Inverted vehicle double main shaft synchronous anti-collision control method and system

Abstract

The application relates to the technical field of numerical control, and discloses an inverted vehicle double-main-shaft synchronous anti-collision control method and system. The method comprises the following steps: periodically sampling the positions, speeds, torques and clamping forces of the primary and secondary main shafts to obtain a state vector, identifying the equivalent mass, available deceleration and identification convergence state by means of motor torque online identification, switching the double-mode criterion of the relative momentum dissipation margin and the momentum transmission deviation margin by triggering the double-edge zero-crossing event of the clamping force, and performing the anti-saturation limiting control of the servo speed loop on the basis of the upper limit of the feeding acceleration according to the results. The application shortens the end-to-end response time required for the double-main-shaft synchronous anti-collision control from criterion over-limit to physical deceleration of the servo actuator.

Description

Technical Field

[0001] This application relates to the field of CNC control technology, and in particular to a method and system for synchronous anti-collision control of dual spindles on an inverted lathe. Background Technology

[0002] An inverted lathe is a type of vertical turning equipment that rotates the spindle of a traditional horizontal lathe by 90 degrees and makes the chuck opening face downwards. Its spindle itself undertakes the integrated action of picking up the workpiece, transferring the machining area, and returning it after machining, without the need for an external loading and unloading robot. When it is further equipped with a main spindle and a secondary spindle and makes the two move in opposite directions in the X and Z directions on a shared bed, it forms a double-spindle inverted lathe. It can complete turning, milling, end face, outer circle, and inner hole machining of both ends of the workpiece in one clamping. In the industry, its process coverage capability is summarized as double pentahedral composite machining. Existing CNC systems typically employ a three-dimensional geometric interference judgment approach for synchronous collision avoidance control of dual-spindle structures. Before system deployment, based on the machine tool's mechanical assembly dimensions, the main spindle, auxiliary spindle, and the chuck, fixture, and workpiece moving with the spindles are abstracted into cuboids or cylinders according to their maximum contours and stored in a static interference table. During machining, the position coordinates of each feed axis in the X and Z directions are collected according to the interpolation cycle. These coordinates are then substituted into the pre-stored envelope equations, and the distance between the two spindle envelopes at the current moment is determined using the separation axis theorem, axis alignment bounding box intersection, or geometric nearest distance algorithm. If the distance is lower than a preset safety threshold, the programmable logic controller adjusts the feed speed ratio of the two axes to cause the two spindles to decelerate collaboratively, the interpolation output in a certain direction is stopped by inter-axis interlocking, or the servo brake is driven by an emergency stop signal. For the workpiece handover process, the main program inserts position synchronization points before and after the handover using M-code synchronization instructions or multi-channel waiting instructions. Subsequent actions are only allowed after both parties have reached the predetermined handover posture and the clamping and releasing signals have been double-confirmed.

[0003] The aforementioned dual-spindle anti-collision control route based on 3D geometric interference judgment and M-code handshake protection has the following shortcomings: First, the interference region model uses the maximum geometric envelope of offline modeling as the judgment boundary, which cannot be updated in real time according to the actual contour after the workpiece material is removed, resulting in a large amount of redundancy in the envelope relative to the actual contour during the finishing stage, forcing the control system to decelerate or stop prematurely; Second, the geometric interference judgment only uses position coordinates as input, making it difficult to identify the risk of chuck side rubbing caused by the phase difference of the C-axis of the main and auxiliary spindle chucks within the interpolation cycle, and even if the rotation dimension is introduced, it is only processed by periodic polling and cannot achieve the same periodic sampling as the position loop; Third, the magnification adjustment and emergency stop response are issued through the upper-level instruction channel of the CNC system, and after parsing by the main program, handshake by the programmable logic controller, and redistribution of servo instructions, the response is not fully implemented. The end-to-end delay is of the same order as the interpolation cycle. The millisecond-level delay makes it risky that the deceleration distance is insufficient to dissipate relative momentum when the relative speed of the two spindles is large or the equivalent mass of the workpiece is large. Fourth, the M-code handshake mechanism in the workpiece handover process only verifies the order of the logic signals. It does not cross-confirm the actual position closed-loop deviation of the main and auxiliary spindles in the handover window, the servo speed following deviation, and the measured value of the chuck cylinder clamping force. When servo following error, jaw wear, or hydraulic pressure fluctuation occurs, there is a risk that both ends report to be in place but the actual position is misaligned, which may cause the workpiece to be squeezed or fall off at the moment of clamping-releasing. Fifth, the anti-collision criterion of the machining section and the handshake protection of the handover section are processed as two independent logics in parallel. The anti-collision criterion must be shielded during the handover window to avoid false triggering of the active approach action, thus forming a protection blind spot during the shielding period. Summary of the Invention

[0004] This application provides a synchronous anti-collision control method and system for dual spindles of an inverted lathe, which solves the problems in the existing anti-collision control of dual spindles of inverted lathes where the static geometric envelope criterion cannot follow the dynamic quality changes of the workpiece and the protection blind zone caused by the shielding switch between the anti-collision criterion of the machining section and the handshake protection of ...

[0005] In a first aspect, this application provides a method for synchronous anti-collision control of dual main shafts on an inverted vehicle, the method comprising: Step S1: Based on the machine tool CNC system, perform synchronous sampling control of the position coordinates, feed speed, motor torque and clamping force of the main spindle and the auxiliary spindle in the same period to obtain a real-time state vector of the dual spindles including the position coordinates, the feed speed, the motor torque and the clamping force; Step S2: Based on the motor torque in the real-time state vector of the dual spindles, perform servo current loop closed-loop online identification control on the equivalent mass and available deceleration of the positive spindle and the secondary spindle, and output the equivalent mass, the available deceleration, and the identification convergence state characterizing whether the online identification control has converged. Step S3: Using the double-sided zero-crossing event of the clamping force in the real-time state vector of the dual spindles as the switching trigger condition, the relative momentum is synthesized by the equivalent mass and the feed speed, and the dissipation required stopping distance is synthesized by the available deceleration and the feed speed. Based on the relative momentum, the dissipation required stopping distance, the position coordinates and the identification convergence state, dual-mode criterion switching control is performed on the relative momentum dissipation margin of the machining section and the momentum transfer deviation margin of the junction section, and the dual-mode anti-collision control state is output. Step S4: Based on the dual-mode anti-collision control state and the identification convergence state, perform servo speed loop anti-saturation limiting control on the upper limit of the feed acceleration of the main spindle and the secondary spindle.

[0006] Secondly, this application provides a dual-spindle synchronous anti-collision control system for an inverted vehicle, the dual-spindle synchronous anti-collision control system comprising: The sampling module is used to perform synchronous sampling control of the position coordinates, feed speed, motor torque and clamping force of the main spindle and the auxiliary spindle based on the CNC system of the machine tool, and obtain a real-time state vector of the dual spindles including the position coordinates, the feed speed, the motor torque and the clamping force; The identification module is used to perform servo current loop closed-loop online identification control on the equivalent mass and available deceleration of the positive spindle and the secondary spindle based on the motor torque in the real-time state vector of the dual spindles, and output the equivalent mass, the available deceleration and the identification convergence state characterizing whether the online identification control has converged. The switching module is used to use the double-sided zero-crossing event of the clamping force in the real-time state vector of the dual spindles as the switching trigger condition, synthesize the relative momentum from the equivalent mass and the feed speed, synthesize the dissipation required stopping distance from the available deceleration and the feed speed, and perform dual-mode criterion switching control on the relative momentum dissipation margin of the machining section and the momentum transfer deviation margin of the handover section based on the relative momentum, the dissipation required stopping distance, the position coordinates and the identification convergence state, and output the dual-mode anti-collision control state. The control module is used to perform servo speed loop anti-saturation limiting control on the upper limit of the feed acceleration of the main spindle and the secondary spindle based on the dual-mode anti-collision control state and the identification convergence state.

[0007] Thirdly, an inverted car dual-spindle synchronous anti-collision control device is provided, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor invokes the instructions in the memory to cause the inverted car dual-spindle synchronous anti-collision control device to execute the above-described inverted car dual-spindle synchronous anti-collision control method.

[0008] Fourthly, a computer-readable storage medium is provided, wherein instructions are stored therein, which, when executed on a computer, cause the computer to perform the above-described inverted car dual-spindle synchronous anti-collision control method.

[0009] In the technical solution provided in this application, the position coordinates, feed speed, motor torque and clamping force of the main spindle and the auxiliary spindle are sampled and controlled synchronously in the same period by the CNC system of the machine tool and assembled into a real-time state vector of the dual spindles. This makes the physical quantities that were originally scattered in the multi-source heterogeneous channels such as the grating ruler, servo driver, chuck cylinder pressure sensor and so on latched and aligned at the same phase point in the same interpolation cycle. This avoids the phase misalignment caused by polling the position and torque quantities according to the channel in the existing CNC anti-collision control, and makes the force-acceleration causal relationship on which the subsequent dynamic judgment depends self-consistent in physical time. Based on this, the servo current loop closed-loop online identification control reverses the instantaneous torque information already present in the current loop torque register and uses it as an identification source for equivalent mass and available deceleration. This allows dynamic changes caused by workpiece material removal, chuck clamping tightness switching, and servo temperature rise to be continuously tracked. As a result, the equivalent mass and available deceleration are transformed from offline calibrated static parameters into online quantities that are updated in real time as the process progresses. The identification convergence state of the accompanying output further provides a reliable basis for downstream processes, enabling downstream judgments to automatically switch to conservative handling at the moment when the identification is not yet stable, rather than making rash decisions based on unreliable identification values. Overall, the uncertainty of dynamic quantities is incorporated into the control loop in a traceable form.

[0010] The arrangement of using the bilateral zero-crossing event of the clamping force in the real-time state vector of the dual spindles as the switching trigger condition allows the two semantically opposite criteria, the relative momentum dissipation margin of the machining section and the momentum transfer deviation margin of the handover section, to share the same physical quantity, momentum, and switch mutually exclusively according to the unforgeable physical event. This avoids the protection blind spot formed by the M-code handshake to shield the anti-collision criteria of the handover window in the existing technology, so that the anti-collision control under the two working conditions of machining and handover is no longer artificially separated into two independent logics. The processing path that synthesizes the relative momentum by equivalent mass and feed rate, and synthesizes the required stopping distance for dissipation by available deceleration and feed rate, carries the dynamic feasibility with a single dimensionless scalar, answering the physical feasibility question of whether it can stop before a collision, rather than the question of whether a collision will occur, which can only be answered by the existing geometric criteria. Thus, it takes into account both response sensitivity and cutting cycle time in the two high-strength coupling working conditions of double pentahedral composite machining and workpiece handover.

[0011] The final output dual-mode anti-collision control state, combined with the identification convergence state, serves as the basis for servo speed loop anti-saturation limiting control. The physical location of the limiting injection point is moved from the upper-level magnification channel of the CNC system to the anti-saturation link of the servo driver speed loop PI regulator. This allows the limiting information carried by the algorithm criterion to directly act on the speed loop output clamping, rather than through a multi-level path involving main program parsing, programmable logic controller handshaking, and servo command redistribution. From a control engineering perspective, this scheme, in the specific application field of dual-spindle synchronous anti-collision for inverted lathes, achieves a substantial difference from existing CNC anti-collision technologies by synergistically contributing three algorithmic features: online identification of equivalent mass, dual-mode switching of momentum dissipation and momentum transfer, and servo-layer anti-saturation limiting injection. This elevates CNC anti-collision control from trajectory monitoring at the geometric kinematic level to feasible clamping at the servo dynamics level. Attached Figure Description

[0012] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a schematic diagram of one embodiment of the inverted car dual-spindle synchronous anti-collision control method in this application; Figure 2 This is a timing diagram for the clamping force bilateral zero-crossing event detection in the embodiments of this application; Figure 3 This is a cumulative distribution map of the dwell time in the three sections in the embodiments of this application. Detailed Implementation

[0014] This application provides a method and system for synchronous anti-collision control of a dual-spindle inverted vehicle. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0015] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the inverted car dual-spindle synchronous anti-collision control method in this application includes: Step S1: Based on the machine tool CNC system, perform synchronous sampling control of the position coordinates, feed speed, motor torque and clamping force of the main spindle and the auxiliary spindle in the same period to obtain the real-time state vector of the dual spindles, which includes position coordinates, feed speed, motor torque and clamping force. Specifically, the main spindle and the auxiliary spindle are arranged in opposite directions along the X and Z axes in the inverted lathe structure and each drives the C-axis rotation. The grating ruler, current loop torque register, and chuck cylinder pressure sensor are physically distributed at different servo nodes on both sides of the machine tool. If traditional CNC polling is used to read the data by channel, the position and torque values ​​will be misaligned by more than several hundred microseconds on the time axis, leading to a systematic deviation when the equivalent mass is calculated by pairing position and torque. The synchronous sampling control of the same period uses the distributed clock of the servo bus as a reference to lock the reading trigger of the four types of heterogeneous data to the same phase point of the same interpolation cycle, making the real-time state vector of the dual spindles, consisting of position coordinates, feed rate, motor torque, and clamping force, inseparable on the time axis.

[0016] Step S2: Based on the motor torque in the real-time state vector of the dual spindles, perform servo current loop closed-loop online identification control on the equivalent mass and available deceleration of the positive spindle and the secondary spindle, and output the equivalent mass, available deceleration, and identification convergence status that characterizes whether the online identification control has converged. Specifically, during the machining process of the inverted lathe, the workpiece quality continuously decreases due to material removal. The chuck clamping and loosening switches cause a step change in the equivalent moment of inertia. Furthermore, the torque margin of the servo motor decays with thermal load. These three types of changes cannot be covered by preset static parameters in CAD. The essence of the servo current loop closed-loop online identification control is to convert the current loop torque register reading into a linear driving force via the lead screw. After deducting the cutting reaction force estimated based on the current spindle load current and the pre-calibrated guide rail friction term, this force is compared with the measured acceleration obtained by the grating ruler to form a driving force-acceleration pair. This pair is then input into a recursive least squares identifier with a forgetting factor to update the equivalent quality cycle by cycle. The identification convergence state is given by the continuous comparison between the identification residual and the preset residual threshold. Convergence is defined when the residual remains below the threshold for several interpolation cycles, allowing subsequent collision avoidance criteria to determine the reliability of the current equivalent quality. If the identification has not yet converged when the workpiece is first gripped or released, the downstream processes are conservatively handled as non-converged branches.

[0017] Step S3: Using the double-sided zero-crossing event of the clamping force in the real-time state vector of the dual spindles as the switching trigger condition, the relative momentum is synthesized by the equivalent mass and the feed rate, and the dissipation required stopping distance is synthesized by the available deceleration and the feed rate. Based on the relative momentum, the dissipation required stopping distance, the position coordinates and the identified convergence state, the dual-mode criterion switching control is performed on the relative momentum dissipation margin of the machining section and the momentum transfer deviation margin of the handover section, and the dual-mode anti-collision control state is output. Specifically, the physical threat of the two spindles approaching each other along the X / Z direction in the machining section depends on "whether the relative momentum can be dissipated to zero within the remaining distance," rather than the geometric distance itself. Therefore, the relative momentum dissipation margin is jointly determined by the relative momentum synthesized by the equivalent mass and feed rate, and the stopping distance required for dissipation synthesized by the available deceleration and feed rate. In the handover section, the two spindles actively approach each other under command to complete the double-sided clamping of the workpiece. Geometric approach is a normal process. At this time, the physical threat of concern shifts to "whether the combined momentum of the coupled system composed of the two spindles and the workpiece converges towards the target momentum," which is borne by the momentum transfer deviation margin. The two types of margins share the same physical quantity of momentum, but they respectively represent two opposite physical processes: dissipation and transfer. Switching cannot rely on logic signals or timing handshakes, otherwise there is a risk of false switching when there are servo following errors or hydraulic pressure fluctuations. Therefore, the double-sided zero-crossing event of clamping force is selected as the switching trigger condition—only when the pressure of the hydraulic cylinders of the positive and negative chucks simultaneously crosses the clamping threshold, which is an unforgeable physical event, is the double-sided coupling relationship established and the criterion mode switched. Identifying the convergence state plays a role in threshold conservatism during switching, narrowing the safe zone when the equivalent quality has not converged.

[0018] Step S4: Based on the dual-mode anti-collision control state and the identification convergence state, perform servo speed loop anti-saturation limiting control on the upper limit of the feed acceleration of the main spindle and the secondary spindle.

[0019] Specifically, the path of the upper limit of feed acceleration determines the time constant of the anti-collision control response. The CNC upper-level scaling channel is parsed by the main program, handshake with the PLC, and redistributed by the servo command, with the end-to-end delay being of the same order as the interpolation cycle. The servo speed loop anti-saturation limiting register is located in the output limiting link of the speed loop PI regulator inside the driver. Writing a value to it takes effect within the driver firmware cycle, making the limiting response delay of the same order as the current loop cycle. The safety zone, warning zone, and irreversible differentiation classification codes carried in the dual-mode anti-collision control state are mapped by the limiting gear and combined with the conservative modulation of the convergence state to determine the reference upper limit value. Then, they are converted into components in the X and Z directions and written to the anti-saturation register of the corresponding servo driver. This ensures that when the warning occurs in the machining section, the sub-spindle has already started physical deceleration within the current interpolation cycle. When the momentum deviation exceeds the limit in the handover section, the two spindles converge momentum synchronously instead of relying on the CNC layer to reissue the scaling command. The selection of the limiting injection point constitutes the key technical contribution that distinguishes this scheme from existing CNC scaling anti-collision schemes.

[0020] In one specific embodiment, step S1 includes: Based on the interpolation cycle of the machine tool CNC system, EtherCAT distributed clock synchronization processing is performed on the servo bus clocks of the main spindle and the auxiliary spindle to obtain a same-cycle sampling trigger signal with a synchronization deviation of no more than 1μs. Based on the same period sampling trigger signal, the X-axis grating ruler and Z-axis grating ruler of the main spindle and the sub-spindle are read simultaneously to obtain the position coordinates. Based on the position coordinates, the first-order differential processing is performed between two adjacent interpolation cycles to obtain the feed rate. Based on the same-period sampling trigger signal, the current loop torque registers of the X-axis servo driver and Z-axis servo driver of the main spindle and the auxiliary spindle are read simultaneously to obtain the motor torque; Based on the same period sampling trigger signal, the chuck cylinder pressure sensors of the main spindle and the auxiliary spindle are read simultaneously to obtain the clamping force. Then, based on the position coordinates, feed speed, motor torque and clamping force, the time stamp of the interpolation period is used to perform simultaneous alignment and combination processing to obtain the real-time status vector of the dual spindles.

[0021] Specifically, EtherCAT distributed clock synchronization processing broadcasts the master clock as a reference source to each slave node at the servo bus physical layer. Each slave node's local clock continuously corrects its phase difference with the master clock through hardware-level timestamp comparison. After tens of milliseconds of initial convergence, clock drift between nodes is suppressed to the order of hundreds of nanoseconds. The resulting synchronous sampling trigger signal jumps simultaneously at all servo nodes connected to the main and secondary spindles. The synchronization deviation is constrained to no more than 1μs, with an upper limit of 0.5% of the interpolation period of 2ms. This ensures that the feed rate obtained from position coordinate differential does not introduce a speed error exceeding 0.5% between adjacent cycles due to sampling misalignment. Simultaneously, this value is higher than the lower limit of the servo current loop period of 62.5μs, avoiding unrealistic requirements on the bus hardware. The difference between the position coordinates read by the grating ruler and adjacent interpolation cycles, divided by the interpolation period length, constitutes the feed rate output of the first-order differential processing. Before differential processing, the original position code track undergoes four-point sliding average preprocessing to suppress high-frequency jitter in the grating signal.

[0022] The current loop torque register stores the instantaneous output torque word obtained by the servo driver from the q-axis current through a torque coefficient in the most recent current loop cycle. This register is updated once at the end of each current loop cycle. When the sampling trigger signal arrives in the same cycle, the current value is directly latched to obtain the motor torque, without going through the intermediate processing path of the driver's main control CPU. This avoids the millisecond-level delay introduced by reading through MODBUS or Profinet protocols. The chuck cylinder pressure sensor outputs the measured oil pressure digital quantity of the cylinder hydraulic chamber. It is linearly converted into the positive spindle side clamping force and the secondary spindle side clamping force according to the pre-calibrated oil pressure-clamping force conversion coefficient. The calibration coefficient includes the product of the effective working area of ​​the piston and the mechanical efficiency of the cylinder. After the four types of heterogeneous data are simultaneously latched at the rising edge of the synchronous sampling trigger signal in the same interpolation cycle, they are written into a fixed-length data structure of the dual spindle real-time state vector according to the predetermined field order. Each field carries a sampling timestamp for downstream links to perform simultaneous alignment and combination verification, so that the position coordinates, feed rate, motor torque and clamping force are physically inseparable to form the same frame of data.

[0023] In one specific embodiment, step S2 includes: Based on the motor torque and preset lead screw pair lead in the real-time state vector of the dual spindles, the driving force conversion processing is performed on the X and Z directions of the main spindle and the secondary spindle respectively to obtain the net driving force in the X direction and the net driving force in the Z direction. Based on the feed rate in the real-time state vector of the dual spindles, a first-order difference processing is performed between adjacent interpolation cycles to obtain the measured acceleration in the X direction and the measured acceleration in the Z direction. The net driving force in the X direction, the net driving force in the Z direction, the measured acceleration in the X direction, and the measured acceleration in the Z direction are input into a recursive least squares identifier with a forgetting factor to perform online identification processing, thereby obtaining the equivalent mass of the positive principal axis and the secondary principal axis. Based on the identification residual of the recursive least squares identifier and the preset residual threshold, a convergence judgment process is performed to obtain the identification convergence state. The torque margin is calculated based on the difference between the rated torque of the servo drive and the motor torque of the main spindle and the secondary spindle. The available deceleration is then calculated based on the torque margin, the lead screw pitch, and the equivalent mass to obtain the available deceleration.

[0024] Specifically, when the motor torque is converted into linear driving force in the X or Z direction via the lead screw, the geometric conversion relationship is followed: driving force equals 2π multiplied by the motor torque and then divided by the lead screw. The lead screw is taken as the actual ball screw lead value assembled on the machine tool; for inverted lathes, 10 mm is commonly used in the X direction and 12 mm in the Z direction. After deducting two terms from the converted linear driving force, the net driving force in the X and Z directions are obtained. The cutting reaction force is estimated based on the spindle load current using the Kienzle cutting force model. The model coefficients are pre-calibrated according to the workpiece material grade and stored in the CNC system parameter table before machining. Specifically, the spindle load current is read by the spindle servo drive current loop and converted into spindle cutting torque using the spindle motor torque coefficient. Cutting torque Current machining radius of the workpiece The ratio yields the tangential cutting force. ,Right now:

[0025] Kienzle's cutting force model expresses the tangential cutting force as:

[0026] in The unit cutting force is defined as the cutting width. =1 mm and undeformed chip thickness = 1 mm tangential cutting force, in Newtons per square millimeter; The cutting width is determined by the depth of cut parameter of the current process, and the unit is millimeters. The thickness of the undeformed chip is determined by the feed rate per revolution. With principal angle The product of the sine values ​​is determined, that is:

[0027] The Kienzle exponent is a dimensionless material constant that reflects the nonlinearity of the cutting force as a function of the thickness of the undeformed chip. Model coefficients. and Before machining, the workpiece material grade is pre-calibrated and stored in the CNC system parameter table. The calibration process is as follows: For the workpiece material grade to be machined, a fixed cutting width is used on the same inverted lathe. 1. Fix the spindle speed and sequentially set no fewer than five different feed rates per revolution. This makes the corresponding undeformed chip thickness Covering the range from finishing to roughing, steady-state cutting is performed for each feed rate, and the corresponding measured tangential cutting force is recorded by the spindle load current. Taking the natural logarithm of both sides of the Kienzle model yields:

[0028] by For independent variable, Performing least squares linear regression on the dependent variable, the regression slope is... Therefore, we can conclude that:

[0029] The regression intercept is Therefore, we can conclude that:

[0030] For typical inverted lathe machining materials, carbon structural steel Take 1500 to 2100 Newtons per square millimeter, Take 0.17 to 0.26, cast iron Take 790 to 1350 Newtons per square millimeter, Take 0.20 to 0.28, aluminum alloy Take 500 to 900 Newtons per square millimeter, Take a value between 0.20 and 0.30. (Calibration result) and The material grade is entered into the CNC system parameter table, and the corresponding coefficient is automatically retrieved based on the current workpiece material grade when the machining program starts. The tangential cutting force is obtained from the Kienzle model. The cutting reaction force is decomposed into X-axis and Z-axis cutting reaction force components based on the geometric projection relationship between the current tool principal cutting edge angle, the workpiece coordinate system, and the bed coordinate system. These components are then deducted from the linear driving force in the corresponding directions.

[0031] The guide rail friction force is calculated as the product of the pre-calibrated guide rail friction coefficient and the normal load. The typical value of the guide rail friction coefficient for rolling guide rails is between 0.008 and 0.015. The measured acceleration in the X direction and the measured acceleration in the Z direction are obtained by subtracting the feed rate between two adjacent interpolation cycles and then dividing by the interpolation cycle length. The feed rate itself has been pre-processed by the four-point sliding average in the previous stage.

[0032] The recursive least squares identifier with a forgetting factor uses the net driving force in the X or Z direction as the observed value, the measured acceleration in the corresponding direction as the regression vector element, and the equivalent mass as the parameter to be identified. The forgetting factor is set to 0.98 so that the identifier assigns major weight to the data of the most recent fifty interpolation cycles and exponentially decays the data of earlier data, thereby tracking the slow drift of equivalent mass caused by workpiece material removal and the step change of equivalent mass caused by chuck release. In each interpolation cycle, the identifier updates the covariance matrix and parameter estimates in a standard recursive form, and outputs the identification residual, which is the absolute value of the difference between the observed value and the regression prediction value. The preset residual threshold is determined based on the statistical upper limit of the identifier residual under no-load conditions. For the inverted spindle system of kilogram-level machines, it is set to 5 Newtons. The identification convergence state is set when the residual is continuously lower than this threshold for 16 consecutive interpolation cycles; otherwise, it is set to non-convergence. Torque margin, the difference between the rated torque on the servo drive nameplate and the current motor torque, reflects the remaining braking torque margin that the servo system can output. The calculation of available deceleration follows the physical causal relationship of converting the torque margin into available braking force by the ratio of 2π to the lead screw, and then dividing by the equivalent mass. The calculation is performed independently for the X and Z axes of the main and auxiliary spindles, ensuring that the available deceleration increases with the decrease in equivalent mass due to workpiece material removal and decreases with the decay of torque margin due to servo temperature rise, thus providing a real-time representation of the maximum physically available deceleration capability.

[0033] In one specific embodiment, step S3 uses the bilateral zero-crossing event of the clamping force in the real-time state vector of the dual spindles as the switching trigger condition, including: Based on the clamping force in the real-time state vector of the dual spindles, the clamping force on the positive spindle side and the clamping force on the secondary spindle side are compared with the preset clamping thresholds to obtain the clamping complete flags on the positive spindle side and the secondary spindle side. Based on the clamping position flag on the positive spindle side and the clamping position flag on the secondary spindle side, edge detection processing is performed on the rising edge of the two flags from non-simultaneous setting to simultaneous setting and the falling edge from simultaneous setting back to either side not set, to obtain the double-sided zero-crossing event. Based on the bilateral zero-crossing event, the steady-state persistence of the bilateral zero-crossing event within N consecutive interpolation periods is subjected to de-jitter filtering to obtain the de-jitter confirmed bilateral zero-crossing event; Based on the rising and falling edges of the double-sided zero-crossing events confirmed by debouncing, the state machine is set to perform the switching between the processing segment identifier and the handover segment identifier to obtain the switching trigger condition.

[0034] Specifically, the preset clamping threshold is determined based on the minimum effective clamping pressure of the chuck cylinders of the primary and secondary spindles. When the primary spindle performs roughing operations, the cutting reaction force is larger, so a higher clamping threshold is used. When the secondary spindle performs finishing and transition operations, the cutting reaction force is smaller, so a lower clamping threshold is used. The primary and secondary spindles are set independently to avoid one side being overly sensitive or the other unresponsive due to a single threshold. Under typical operating conditions of the inverted lathe, the clamping threshold for the primary spindle is on the order of 500 Newtons, and for the secondary spindle, it is on the order of 400 Newtons. Specific values ​​are determined through calibration tests during machine tool assembly and entered into the CNC system parameter table. When the clamping force on the primary spindle is higher than the clamping threshold, the clamping position on the primary spindle is set to one; otherwise, it is set to zero. The secondary spindle is determined independently using the same logic. Both flags being one indicates that the workpiece has been effectively clamped by both chucks simultaneously, constituting the physical premise for momentum coupling between the two spindles. The term "double-sided zero-crossing event" refers to the zero-crossing transition event of the logic values ​​of two flag bits, rather than the clamping force itself crossing the zero point. The rising edge corresponds to the instant when the two flag bits change from not being simultaneously one to being simultaneously one, indicating that the workpiece has just been successfully clamped by both sides and the processing section has switched to the handover section. The falling edge corresponds to the instant when the two flag bits return from being simultaneously one to being zero on either side, indicating that the bilateral coupling relationship has broken and the handover section has switched to the processing section.

[0035] Edge detection performs an XOR comparison between the result of the logical AND operation of the two flag bits and the result of the logical AND operation of the previous cycle in each interpolation cycle to determine whether a transition has occurred in the current cycle and the direction of the transition. During the hydraulic pressure build-up phase of the inverted lathe's hydraulic clamping system, hydraulic pressure fluctuations exist, and the measured clamping force may fluctuate around the clamping threshold, causing frequent transitions of the two flag bits. De-jitter filtering accumulates and confirms the steady-state of the edge detection result over N consecutive interpolation cycles. A transition is only recorded as a de-jitter-confirmed double-sided zero-crossing event if the result of the logical AND operation of the two flag bits remains unchanged for N consecutive interpolation cycles after the transition. N is an integer between 4 and 8, corresponding to a de-jitter window length of 4 to 16 milliseconds, avoiding millisecond-level fluctuations during the hydraulic pressure build-up phase without introducing significant switching delays. The state machine includes two mutually exclusive states: a processing segment and a handover segment. Initialization sets it to the processing segment, with the processing segment flag and the handover segment flag being inverses of each other, and only one being true at any given time. The rising edge of a double-sided zero-crossing event, confirmed by debouncing, triggers the state machine to transition from the processing section to the handover section, setting the processing section identifier to zero and the handover section identifier to one. The falling edge triggers the state machine to revert from the handover section to the processing section, setting the processing section identifier to one and the handover section identifier to zero. The current state of the state machine outputs a switching trigger condition. Downstream processes use this switching trigger condition to make a binary choice between relative momentum dissipation margin and momentum transfer deviation margin, ensuring that the criteria for the processing section and the handover section are mutually exclusive based on an unforgeable double-sided clamping physical event, preventing logical gaps where both types of criteria are simultaneously effective or ineffective.

[0036] Figure 2This is a timing diagram for the clamping force bilateral zero-crossing event detection in the embodiments of this application. Figure 2 The illustration shows the process of establishing the clamping force on the main spindle side and the clamping force on the secondary spindle side over time in the embodiments of this application, the edge detection results of the clamping position flags on both sides changing from low level to high level, and the timing relationship of the state change of the double-sided zero-crossing event output after continuous interpolation cycle de-jitter filtering, in order to demonstrate the implementation process of using the double-sided zero-crossing event of clamping force as the switching trigger condition for the processing section and the handover section.

[0037] In one specific embodiment, step S3 involves synthesizing the relative momentum from the equivalent mass and feed rate, synthesizing the required stopping distance for dissipation from the available deceleration and feed rate, and performing calculation processing on the relative momentum dissipation margin of the machining section, including: Based on the feed rate in the real-time state vector of the dual spindles, vector difference processing is performed on the feed rate of the positive spindle and the feed rate of the secondary spindle to obtain the relative feed rate vector. Based on the equivalent mass and the relative feed rate vector, the equivalent mass of the positive spindle and the equivalent mass of the secondary spindle are subjected to a reduced mass synthesis process, and the result of the reduced mass synthesis is subjected to a vector product process with the relative feed rate vector to obtain the relative momentum; Based on the relative feed rate vector and available deceleration, the approach component of the relative feed rate vector in the direction of the relative position of the two main axes is projected, and the ratio of the square of the approach component to twice the system-level available deceleration is calculated to obtain the stopping distance required for dissipation. Based on the position coordinates in the real-time state vector of the dual spindles, the difference between the geometric distance between the primary spindle and the secondary spindle and the preset envelope radius is calculated to obtain the remaining safety distance. Based on the ratio of the required stopping distance for dissipation to the remaining safety distance, a margin calculation is performed to obtain the relative momentum dissipation margin of the machining section.

[0038] Specifically, the feed velocity vectors of the primary and secondary principal axes are assembled into two-dimensional planar vectors by the X-axis and Z-axis feed velocities of the corresponding principal axes. Vector difference processing subtracts the primary principal axis feed velocity vector from the secondary principal axis feed velocity vector to obtain the relative feed velocity vector. The X-axis and Z-axis components of this vector represent the motion rates of the secondary principal axis relative to the primary principal axis in the two feed directions, respectively. The reduced mass synthesis process multiplies the equivalent mass of the primary and secondary principal axes and divides it by the sum of the two. The resulting reduced mass is the equivalent mass used in classical mechanics when a two-particle coupled system is treated as a single-particle system, reflecting the effective inertia of the two principal axes as a coupled system participating in momentum exchange in the relative motion dimension. The vector product of the reduced mass and the relative feed velocity vector yields the relative momentum, which is a vector quantity, oriented along the relative feed velocity vector, and its magnitude is the product of the magnitudes of the reduced mass and the relative feed velocity vector.

[0039] The positional coordinate difference between the primary and secondary spindles in the bed coordinate system constitutes the relative position vector of the two spindles, pointing from the primary spindle to the secondary spindle. The projection of the relative feed rate vector onto the direction of the relative position vector of the two spindles is the approach component. A positive projection indicates that the secondary spindle is approaching the primary spindle with the approach component value. A negative or zero projection indicates that the two spindles are moving away from each other or maintaining an equal distance. Only when the projection is positive is there a risk of physical collision. Therefore, when the approach component is negative or zero, the required stopping distance for dissipation is directly set to zero and subsequent calculations are skipped. The system-level available deceleration is the smaller of the available deceleration of the primary spindle and the available deceleration of the secondary spindle, reflecting the upper limit of system-level braking limited by the weaker deceleration capability of the two spindles according to the conservative principle. The square of the approach component divided by twice the system-level available deceleration is derived from the kinematic equation of uniform deceleration. The resulting required stopping distance for dissipation is the relative displacement required for the two spindles to come to a complete stop under the current approach component. The preset envelope radius is obtained by adding the maximum outer diameter of the main spindle chuck, the maximum outer diameter of the auxiliary spindle chuck, and the maximum radial dimension of the clamped workpiece radially. For typical inverted lathe operation, this is taken as an 80 mm value, and the calibrated value is written into the CNC system parameter table. The remaining safety clearance is obtained by subtracting the preset envelope radius from the modulus of the difference in the position coordinates of the main and auxiliary spindles, representing the currently available distance between the geometric envelopes of the two spindles. The stopping distance required for dissipation divided by the remaining safety clearance yields the relative momentum dissipation margin of the machining section. A ratio less than one indicates that the current approach trend can be completely dissipated physically before the geometric envelopes of the two spindles collide. A ratio greater than or equal to one indicates that the process has entered a physically irreversible phase. This margin serves as a single dimensionless scalar criterion for the coupling of relative kinematic and relative dynamic quantities.

[0040] In one specific embodiment, step S3 involves calculating the momentum transfer deviation margin of the transfer section and performing dual-mode criterion switching control between the processing section and the transfer section, including: Based on the equivalent mass of the main spindle, the equivalent mass of the secondary spindle, and the feed rate in the real-time state vector of the two main spindles, the momentum vectors of the main spindle and the secondary spindle are summed to obtain the combined momentum vector of the two main spindles. The momentum transfer deviation is calculated based on the difference between the resultant momentum vector of the two main shafts and the preset target momentum vector, and the margin is calculated based on the ratio of the momentum transfer deviation to the preset momentum transfer tolerance to obtain the momentum transfer deviation margin of the junction section. Based on the switching triggering condition, a two-choice gating process is performed on the relative momentum dissipation margin of the processing section and the momentum transfer deviation margin of the handover section. Then, conservative modulation processing is performed on the safety threshold of the selected gating margin according to the identified convergence state to obtain the current effective margin and the current safety threshold. Based on the current effective margin and the current safety threshold, an over-limit comparison is performed. The over-limit results are then classified into safe zones, warning zones, and irreversible zones and coded accordingly to obtain a dual-mode collision avoidance control state.

[0041] Specifically, the positive spindle momentum vector is obtained by multiplying the equivalent mass of the positive spindle by its feed rate vector, and the secondary spindle momentum vector is obtained by multiplying the equivalent mass of the secondary spindle by its feed rate vector in the same form. Both are absolute momentum vectors of a single spindle relative to the bed coordinate system, and are physically dual to relative momentum. The resultant momentum vector of the two spindles is obtained by adding the positive spindle momentum vector and the secondary spindle momentum vector in the bed coordinate system according to their components, reflecting the resultant momentum of the coupled system consisting of the two spindles and the workpiece they hold in the bed coordinate system. The preset target momentum vector is set according to the type of handover process. For static handover of disc-type workpieces, a zero vector is used to indicate that the system should be stationary after the handover. For handover with flipping or online measurement, a non-zero vector along the Z-axis is used to indicate that the system should maintain a predetermined uniform flipping or uniform transition state after the handover. The specific value of the target momentum vector is issued as a macro variable in the machining program according to the process number. The magnitude of the difference between the resultant momentum vector of the two main shafts and the preset target momentum vector is the momentum transfer deviation, which characterizes the degree to which the resultant momentum of the current coupled system deviates from the target. The preset momentum transfer tolerance is determined based on the maximum allowable momentum residual at the final state of the handover. For a kilogram-level inverted lathe main shaft system, it is taken as being on the order of 2 Nm / s, and the calibration value is written into the CNC system parameter table. The momentum transfer deviation divided by the preset momentum transfer tolerance yields the momentum transfer deviation margin of the handover section. A ratio less than 1 indicates that the resultant momentum is converging towards the target momentum and the handover is under control. A ratio greater than or equal to 1 indicates that the resultant momentum deviates from the target beyond the allowable range and the handover is unstable.

[0042] The dual-criteria gating process uses the current values ​​of the processing segment identifier and the handover segment identifier in the switching trigger condition for mutually exclusive gating. If the processing segment identifier is active, the relative momentum dissipation margin of the processing segment is selected as the current effective margin, and the corresponding safety threshold of the processing segment is used as the current safety threshold. If the handover segment identifier is active, the momentum transfer deviation margin of the handover segment is selected as the current effective margin, and the corresponding safety threshold of the handover segment is used as the current safety threshold. Only one of the two types of margins participates in the downstream limit comparison at any given time, eliminating the logical conflict of parallel triggering of the two criteria. The conservative modulation process multiplicatively lowers the current safety threshold based on the identified convergence state. If the identified convergence state is converged, the modulation coefficient is set to 1, which maintains the nominal safety threshold. If the identified convergence state is not converged, the modulation coefficient is set to a value between 0.7 and 0.8, tightening the safety threshold more strictly. Thus, in the period when the equivalent quality has not yet been reliably confirmed, the collision avoidance control operates with an earlier intervention and conservative attitude. The over-limit comparison process continuously compares the current effective margin with the current safety threshold. When the current effective margin is lower than the current safety threshold multiplied by 0.7, it is coded as a safe zone; when it is between the current safety threshold multiplied by 0.7 and the current safety threshold, it is coded as a warning zone; and when it reaches or exceeds the current safety threshold, it is coded as an irreversible zone. The three-level hierarchical coding and the segment identifier in the switching trigger condition are assembled into a dual-mode anti-collision control state. The downstream link uses this coding to perform differentiated level mapping in the servo speed loop anti-saturation limiting control. The safe zone maintains the nominal limit, the warning zone attenuates smoothly inversely proportionally, and the irreversible zone is forcibly injected with full-biased deceleration.

[0043] Figure 3 This is a cumulative distribution map of the dwell time in the three sections in the embodiments of this application. Figure 3 The cumulative probability distribution of dwell time when the dual-mode collision avoidance control state is in the safe zone, warning zone, and irreversible zone in the embodiments of this application is shown. The horizontal axis uses a logarithmic scale to show the distribution range of dwell time spanning multiple orders of magnitude in the millisecond range, and the vertical axis shows the cumulative probability of samples below the corresponding dwell time, which is used to reflect the time distribution characteristics of the control state in each segment after the dual-mode criterion switching control is combined with the servo speed loop anti-saturation limiting control.

[0044] In one specific embodiment, step S4 includes: Based on the hierarchical coding results in the dual-mode collision avoidance control state, the limiting gear mapping process is performed on the safe zone, the warning zone and the irreversible zone respectively to obtain the benchmark limiting gear corresponding to the current segment. Based on the identified convergence state, conservative modulation processing is performed on the reference limiting range to obtain the upper limit value of the feed acceleration modulated by the convergence state. Based on the upper limit of the feed acceleration modulated by the convergence state and the current effective margin, the upper limit of acceleration is allocated in the X and Z directions of the main spindle and the sub-spindle respectively to obtain the upper limit components of the X-direction acceleration and the upper limit components of the Z-direction acceleration. The upper limit components of X-axis acceleration and Z-axis acceleration are written into the speed loop anti-saturation limiting registers of the X-axis servo driver and Z-axis servo driver of the main spindle and sub-spindle respectively, and servo speed loop anti-saturation limiting control is performed on the upper limit of feed acceleration of the main spindle and sub-spindle.

[0045] Specifically, the limiting gear mapping process performs piecewise function mapping between the three gears based on the hierarchical coding results. The baseline limiting gear corresponding to the safe zone code takes the nominal value of the system-level available deceleration, i.e., no additional restrictions are imposed. The baseline limiting gear corresponding to the warning zone code is inversely proportionally and smoothly attenuated according to the ratio of the current effective margin to the current safety threshold. The attenuation function is in the form of subtracting the ratio and then squaring it, so that the baseline limiting gear is close to the nominal value at the entrance of the warning zone and approaches zero at the exit of the warning zone, avoiding a step in the limiting value within the warning zone. The baseline limiting gear corresponding to the irreversible zone code takes the negative value of the system-level available deceleration, i.e., a forced injection of full-bias deceleration is equivalent to servo-level emergency braking. The conservative modulation processing in this step targets the reference limiting position itself rather than the safety threshold. When the convergence state is not converged, the reference limiting position is multiplied by a convergence modulation coefficient between 0.7 and 0.8 to further tighten the upper limit of feed acceleration within the period when the equivalent quality has not yet been reliably confirmed, thus obtaining the upper limit of feed acceleration modulated by the convergence state.

[0046] The upper limit of feed acceleration modulated by the convergence state is a system-level scalar, which needs to be allocated to the X and Z directions of the primary and secondary principal axes according to the directional information carried by the current effective margin. The upper limit of acceleration allocation is determined by the direction cosine of the relative position vector in the X and Z directions. When the absolute value of the X component of the relative position vector is larger than the modulus, the X component of the upper limit of acceleration is allocated a larger proportion, and the Z component is allocated a smaller proportion, and vice versa. This ensures that the limiting intensity is applied directionally along the actual direction of the physical threat rather than being isotropically distributed. The primary and secondary principal axes are further split according to the inverse proportion of the equivalent mass of the primary principal axis and the equivalent mass of the secondary principal axis. The side with the smaller equivalent mass is allocated a smaller limiting value and thus bears a greater deceleration burden, which is consistent with the physical intuition of impulse allocation. The speed loop anti-saturation limiting register is located in the limiting stage at the output of the PI regulator in the servo drive's speed loop. It clamps the upper and lower bounds of the current command output by the PI regulator in each current loop cycle. After writing the upper limit component of the X-axis acceleration, this component is converted into the corresponding upper limit of current by the servo drive and directly clamps the speed loop output. The clamping action takes effect within the driver firmware cycle, and the end-to-end delay is on the order of microseconds, the same as the current loop cycle. Compared to the millisecond-level delay path of CNC upper-level magnification commands through main program parsing, PLC handshaking, and servo command redistribution, the servo speed loop anti-saturation limiting control completes the upper limit of feed acceleration within the driver. This allows the physical deceleration response of the two spindles to begin within the current interpolation cycle in two working conditions: machining section warning and momentum deviation exceeding the limit in the handover section, instead of waiting for the next interpolation cycle. The selection of the physical location of the limiting injection point is the fundamental reason for the compression of the anti-collision response time constant in this scheme.

[0047] The above describes the inverted car dual-spindle synchronous anti-collision control method in the embodiments of this application. The following describes the inverted car dual-spindle synchronous anti-collision control system in the embodiments of this application. One embodiment of the inverted car dual-spindle synchronous anti-collision control system in the embodiments of this application includes: The sampling module is used to perform synchronous sampling control of the position coordinates, feed speed, motor torque and clamping force of the main spindle and the auxiliary spindle based on the CNC system of the machine tool, and obtain a real-time state vector of the dual spindles including the position coordinates, the feed speed, the motor torque and the clamping force; The identification module is used to perform servo current loop closed-loop online identification control on the equivalent mass and available deceleration of the positive spindle and the secondary spindle based on the motor torque in the real-time state vector of the dual spindles, and output the equivalent mass, the available deceleration and the identification convergence state characterizing whether the online identification control has converged. The switching module is used to use the double-sided zero-crossing event of the clamping force in the real-time state vector of the dual spindles as the switching trigger condition, synthesize the relative momentum from the equivalent mass and the feed speed, synthesize the dissipation required stopping distance from the available deceleration and the feed speed, and perform dual-mode criterion switching control on the relative momentum dissipation margin of the machining section and the momentum transfer deviation margin of the handover section based on the relative momentum, the dissipation required stopping distance, the position coordinates and the identification convergence state, and output the dual-mode anti-collision control state. The control module is used to perform servo speed loop anti-saturation limiting control on the upper limit of the feed acceleration of the main spindle and the secondary spindle based on the dual-mode anti-collision control state and the identification convergence state.

[0048] This invention also provides a dual-spindle synchronous anti-collision control device for an inverted lathe, which can be a server. The device includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor, designed as a computer, provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores the data corresponding to this embodiment. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements the above-described method.

[0049] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the inverted car dual-spindle synchronous anti-collision control method.

[0050] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0051] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an inverted car dual-spindle synchronous anti-collision control device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0052] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for synchronous anti-collision control of dual main shafts on an inverted lathe, characterized in that, The method includes: Step S1: Based on the machine tool CNC system, perform synchronous sampling control of the position coordinates, feed speed, motor torque and clamping force of the main spindle and the auxiliary spindle in the same period to obtain a real-time state vector of the dual spindles including the position coordinates, the feed speed, the motor torque and the clamping force; Step S2: Based on the motor torque in the real-time state vector of the dual spindles, perform servo current loop closed-loop online identification control on the equivalent mass and available deceleration of the positive spindle and the secondary spindle, and output the equivalent mass, the available deceleration, and the identification convergence state characterizing whether the online identification control has converged. Step S3: Using the double-sided zero-crossing event of the clamping force in the real-time state vector of the dual spindles as the switching trigger condition, the relative momentum is synthesized by the equivalent mass and the feed speed, and the dissipation required stopping distance is synthesized by the available deceleration and the feed speed. Based on the relative momentum, the dissipation required stopping distance, the position coordinates and the identification convergence state, dual-mode criterion switching control is performed on the relative momentum dissipation margin of the machining section and the momentum transfer deviation margin of the junction section, and the dual-mode anti-collision control state is output. Step S4: Based on the dual-mode anti-collision control state and the identification convergence state, perform servo speed loop anti-saturation limiting control on the upper limit of the feed acceleration of the main spindle and the secondary spindle.

2. The inverted car dual-spindle synchronous anti-collision control method according to claim 1, characterized in that, Step S1 includes: Based on the interpolation cycle of the machine tool CNC system, EtherCAT distributed clock synchronization processing is performed on the servo bus clocks of the main spindle and the auxiliary spindle to obtain a same-cycle sampling trigger signal with a synchronization deviation of no more than 1μs. According to the same-period sampling trigger signal, the X-axis grating ruler and Z-axis grating ruler of the main spindle and the sub-spindle are simultaneously read to obtain the position coordinates, and the first-order difference processing is performed between two adjacent interpolation cycles based on the position coordinates to obtain the feed rate. Based on the same-period sampling trigger signal, the current loop torque registers of the X-axis servo driver and Z-axis servo driver of the main spindle and the secondary spindle are simultaneously read to obtain the motor torque; Based on the same-period sampling trigger signal, the chuck cylinder pressure sensors of the main spindle and the auxiliary spindle are simultaneously read to obtain the clamping force. Based on the position coordinates, the feed speed, the motor torque and the clamping force, the timestamp of the interpolation cycle is simultaneously aligned and combined to obtain the real-time state vector of the dual spindles.

3. The inverted car dual-spindle synchronous anti-collision control method according to claim 2, characterized in that, Step S2 includes: Based on the motor torque and preset lead screw pair lead in the real-time state vector of the dual spindles, the driving force conversion processing is performed on the X and Z directions of the positive spindle and the secondary spindle respectively to obtain the net driving force in the X direction and the net driving force in the Z direction. Based on the feed rate in the real-time state vector of the dual spindles, a first-order difference processing is performed between adjacent interpolation cycles to obtain the measured acceleration in the X direction and the measured acceleration in the Z direction. The net driving force in the X direction, the net driving force in the Z direction, the measured acceleration in the X direction, and the measured acceleration in the Z direction are input into a recursive least squares identifier with a forgetting factor to perform online identification processing, thereby obtaining the equivalent mass of the positive principal axis and the secondary principal axis. Based on the identification residual of the recursive least squares identifier and a preset residual threshold, a convergence determination process is performed to obtain the identification convergence state. The torque margin is calculated based on the difference between the rated torque of the servo drive of the main spindle and the auxiliary spindle and the torque of the motor. The available deceleration is then calculated based on the torque margin, the lead screw of the lead screw pair and the equivalent mass to obtain the available deceleration.

4. The inverted car dual-spindle synchronous anti-collision control method according to claim 3, characterized in that, In step S3, the bilateral zero-crossing event of the clamping force in the real-time state vector of the dual spindles is used as the switching trigger condition, including: Based on the clamping force in the real-time state vector of the dual spindles, the clamping force on the positive spindle side and the clamping force on the secondary spindle side are compared with a preset clamping threshold to obtain the clamping position flag on the positive spindle side and the clamping position flag on the secondary spindle side. Based on the clamping position flag on the positive spindle side and the clamping position flag on the secondary spindle side, edge detection processing is performed on the rising edge of the two flags from non-simultaneous setting to simultaneous setting and the falling edge from simultaneous setting back to either side not set, to obtain a double-sided zero-crossing event. Based on the bilateral zero-crossing event, the steady-state continuous state of the bilateral zero-crossing event within N consecutive interpolation cycles is subjected to de-jitter filtering to obtain the de-jitter confirmed bilateral zero-crossing event; Based on the rising and falling edges of the double-sided zero-crossing events confirmed by the debouncing, the state machine is set to perform the switching between the processing segment identifier and the handover segment identifier to obtain the switching trigger condition.

5. The inverted car dual-spindle synchronous anti-collision control method according to claim 4, characterized in that, In step S3, the relative momentum is synthesized from the equivalent mass and the feed rate, and the required stopping distance for dissipation is synthesized from the available deceleration and the feed rate. The relative momentum dissipation margin of the machining section is then calculated, including: Based on the feed rate in the real-time state vector of the dual spindles, vector difference processing is performed on the feed rate of the positive spindle and the feed rate of the secondary spindle to obtain the relative feed rate vector. Based on the equivalent mass and the relative feed rate vector, a reduced mass synthesis process is performed on the equivalent mass of the positive spindle and the equivalent mass of the secondary spindle, and the reduced mass synthesis result is multiplied by the relative feed rate vector to obtain the relative momentum; Based on the relative feed rate vector and the available deceleration, the approach component of the relative feed rate vector in the direction of the relative position of the two main axes is projected, and the ratio of the square of the approach component to twice the system-level available deceleration is calculated to obtain the required stopping distance for dissipation. Based on the position coordinates in the real-time state vector of the dual spindles, the difference between the geometric distance between the positive spindle and the secondary spindle and the preset envelope radius is calculated to obtain the remaining safety distance. Based on the ratio of the required dissipation stopping distance to the remaining safety distance, a margin calculation is performed to obtain the relative momentum dissipation margin of the processing section.

6. The inverted car dual-spindle synchronous anti-collision control method according to claim 5, characterized in that, In step S3, the momentum transfer deviation margin of the transfer section is calculated, and dual-mode criterion switching control is performed between the processing section and the transfer section, including: Based on the equivalent mass of the positive spindle, the equivalent mass of the secondary spindle, and the feed rate in the real-time state vector of the dual spindles, the momentum vector of the positive spindle and the momentum vector of the secondary spindle are summed to obtain the combined momentum vector of the two spindles. The momentum transfer deviation is calculated based on the difference between the sum of the momentum vectors of the two main axes and the preset target momentum vector, and the momentum transfer deviation is calculated based on the ratio of the momentum transfer deviation to the preset momentum transfer tolerance to obtain the momentum transfer deviation margin of the junction segment. Based on the switching triggering condition, a two-choice criterion gating process is performed on the relative momentum dissipation margin of the processing section and the momentum transfer deviation margin of the handover section, and a conservative modulation process is performed on the safety threshold of the selected gating margin according to the identification convergence state to obtain the current effective margin and the current safety threshold. The current effective margin and the current safety threshold are compared to perform an over-limit comparison. The over-limit result is then classified into a safety zone, a warning zone, and an irreversible zone, and a hierarchical coding process is performed to obtain the dual-mode collision avoidance control state.

7. The inverted car dual-spindle synchronous anti-collision control method according to claim 1, characterized in that, Step S4 includes: Based on the hierarchical coding results in the dual-mode collision avoidance control state, the limiting gear mapping process is performed on the safe zone, the warning zone and the irreversible zone respectively to obtain the benchmark limiting gear corresponding to the current segment. Based on the identified convergence state, a conservative modulation process is performed on the reference limiting position to obtain the upper limit value of the feed acceleration modulated by the convergence state. Based on the feed acceleration upper limit value modulated by the convergence state and the current effective margin, the X and Z directions of the positive spindle and the sub-spindle are respectively processed to obtain the X-axis acceleration upper limit component and the Z-axis acceleration upper limit component. The upper limit components of the X-axis acceleration and the upper limit components of the Z-axis acceleration are written into the speed loop anti-saturation limiting registers of the X-axis servo driver and the Z-axis servo driver of the main spindle and the sub-spindle, respectively, and the servo speed loop anti-saturation limiting control is performed on the upper limit of the feed acceleration of the main spindle and the sub-spindle.

8. A dual-spindle synchronous anti-collision control system for an inverted vehicle, characterized in that, For implementing the inverted car dual-spindle synchronous anti-collision control method as described in any one of claims 1-7, the inverted car dual-spindle synchronous anti-collision control system comprises: The sampling module is used to perform synchronous sampling control of the position coordinates, feed speed, motor torque and clamping force of the main spindle and the auxiliary spindle based on the CNC system of the machine tool, and obtain a real-time state vector of the dual spindles including the position coordinates, the feed speed, the motor torque and the clamping force; The identification module is used to perform servo current loop closed-loop online identification control on the equivalent mass and available deceleration of the positive spindle and the secondary spindle based on the motor torque in the real-time state vector of the dual spindles, and output the equivalent mass, the available deceleration and the identification convergence state characterizing whether the online identification control has converged. The switching module is used to use the double-sided zero-crossing event of the clamping force in the real-time state vector of the dual spindles as the switching trigger condition, synthesize the relative momentum from the equivalent mass and the feed speed, synthesize the dissipation required stopping distance from the available deceleration and the feed speed, and perform dual-mode criterion switching control on the relative momentum dissipation margin of the machining section and the momentum transfer deviation margin of the handover section based on the relative momentum, the dissipation required stopping distance, the position coordinates and the identification convergence state, and output the dual-mode anti-collision control state. The control module is used to perform servo speed loop anti-saturation limiting control on the upper limit of the feed acceleration of the main spindle and the secondary spindle based on the dual-mode anti-collision control state and the identification convergence state.

9. A dual-spindle synchronous anti-collision control device for an inverted lathe, characterized in that, The system includes a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the computer program to implement the inverted car dual-spindle synchronous anti-collision control method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, it causes the processor to execute the inverted car dual-spindle synchronous anti-collision control method as described in any one of claims 1 to 7.

Images