Multi-station drawing press stamping force self-adaptive control system and method

Abstract

The present application relates to the technical field of punch press pressure control, in particular to a multi-station stretch drawing press stamping force adaptive control system and method, which utilizes the non-stamping idle window of slider return to online identify the actual kinematic viscosity of hydraulic oil through the fixed restrictor bypass branch, maps the internal leakage compensation and control current bias of electro-hydraulic proportional pressure valve according to the current viscosity, and performs partition interval feedforward correction on the blank holder force and main stamping force of each drawing station, which realizes stamping force adaptive adjustment in cooperation with phased gain closed-loop fine tuning and multi-station independent hydraulic pad driving, can adapt to different working temperatures, oil aging and valve wear conditions, and improves the stamping force control stability and the batch quality consistency of drawn parts.

Description

Technical Field

[0001] This invention relates to the field of pressure control technology for stamping equipment, and more specifically, to an adaptive control system and method for the stamping pressure of a multi-station stretching press. Background Technology

[0002] Multi-station stretching presses are commonly used equipment in the field of sheet metal stamping. They rely on multiple stations to sequentially complete blanking, stretching, shaping, and trimming processes, enabling high-efficiency mass production of complex shell parts. Currently, the stamping pressure control of multi-station stretching presses generally adopts an electro-hydraulic proportional pressure valve with a fixed-parameter closed-loop control scheme. Under standard oil temperature and stable hydraulic performance, this scheme can achieve good pressure control accuracy, meeting the forming requirements of ordinary precision parts. It has been widely used in industrial production. As the pressure transmission medium of the electro-hydraulic control system, the kinematic viscosity of the hydraulic oil changes with the operating temperature. Oil oxidation and aging, as well as the alteration caused by mixing different grades of oil, all contribute to the problem. Equipment typically has its control parameters calibrated at the factory based on valve flow gain and internal leakage characteristics at a standard oil temperature of 40 degrees Celsius. However, in actual production, these control parameters are fixed and not dynamically adjusted according to the actual oil viscosity. Changes in oil viscosity directly alter the valve flow gain of the electro-hydraulic proportional pressure valve, the internal leakage at the valve core-sleeve mating surface, and the hydraulic system's pressure build-up response speed. Under fixed-parameter control, varying degrees of pressure build-up lag, natural pressure drop during the pressure holding phase, inconsistent control accuracy across different pressure ranges, and pressure imbalances between multiple stations can occur. These phenomena can lead to unstable material flow velocity during the stretching process, resulting in quality problems such as wrinkling, cracking, excessive springback, and poor dimensional consistency in the workpiece. Some existing improvement solutions in the industry involve adding an oil temperature sensor to the hydraulic station to indirectly estimate the oil viscosity and perform parameter compensation based on empirical formulas for oil temperature. However, since oil temperature only reflects the overall average temperature of the oil, it is difficult to accurately reflect the actual viscosity of the oil at the valve port, and it is also difficult to compensate for the characteristic drift caused by long-term wear of valve components. The compensation accuracy is limited. Other solutions involve adding an independent viscosity sensor to the hydraulic circuit to directly measure the viscosity. These solutions increase hardware costs and maintenance points, and the viscosity sensor is usually installed in the return oil line. The measured value has a time lag and spatial difference from the actual working viscosity at the valve port, making it difficult to meet the real-time requirements of millisecond-level control in the stamping process. Existing control solutions experience varying degrees of fluctuation in stamping pressure control accuracy throughout the entire temperature range from cold start to continuous working thermal equilibrium, and throughout the entire life cycle of the oil from new oil to aged oil replacement. It is difficult to reliably guarantee the quality consistency of high-precision complex stretched parts. To solve this problem, we provide an adaptive control system and method for stamping pressure of a multi-station stretching press. Summary of the Invention

[0003] The purpose of this invention is to provide an adaptive control system and method for the punching pressure of a multi-station stretching punch press, so as to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, an adaptive control system for the punching force of a multi-station stretching punch press is provided, including: The stamping condition timing identification unit is used to collect the slider-crank angle signal in real time, identify the non-stamping idle window during the slider return phase based on the slider-crank angle signal, and output the identification trigger signal. The online identification unit for actual hydraulic viscosity receives the identification trigger signal and controls the opening of the master cylinder unloading valve to cut off the pressure path of the main actuator. It controls the bypass solenoid valve to open the bypass branch where the fixed throttle is located, inputs a stepped identification current sequence to the electro-hydraulic proportional pressure valve, collects multiple sets of pressure difference values ​​at different steady-state flow rates across the fixed throttle, removes the inherent pressure pulsation component of the hydraulic pump, and then inverts to obtain the actual kinematic viscosity value of the current hydraulic oil. After the identification is completed, the bypass solenoid valve is immediately closed. The valve control characteristic offset mapping unit is used to compare the actual kinematic viscosity value with the reference viscosity value obtained and stored by the same identification process at the standard working oil temperature during the equipment's factory commissioning stage. It queries the pre-stored correspondence between viscosity and valve orifice flow gain to obtain the internal leakage compensation amount and control current bias amount of the electro-hydraulic proportional pressure valve at the current viscosity. The pressure setpoint feedforward correction unit is used to superimpose the blank holder force setpoint of each stretching station with the main punch pressure setpoint and the corresponding control current bias. At the same time, it corrects the current feedforward during the pressure build-up stage and the current maintenance during the pressure holding stage based on the internal leakage compensation, and generates the corrected pressure setpoint. The stamping pressure closed-loop fine-tuning unit is used to collect the actual pressure value of the die holder during the stamping stroke, compare it with the corrected pressure set value, and output the adjustment control quantity to drive the hydraulic cushion actuator of each station to complete the adaptive adjustment of the stamping pressure.

[0005] The second objective of this invention is to provide a method for implementing an adaptive control system for the punching force of a multi-station stretching punch press, comprising the following steps: S1. Real-time acquisition of slider-crank angle signal, identification of non-stamping idle window during slider return phase based on slider-crank angle signal, and outputting identification trigger signal; S2. After receiving the identification trigger signal, control the main cylinder unloading valve to open to cut off the pressure path of the main actuator, control the bypass solenoid valve to open the bypass branch where the fixed throttle is located, input a step-type identification current sequence to the electro-hydraulic proportional pressure valve, collect multiple sets of pressure difference values ​​at both ends of the fixed throttle under different steady-state flow rates, remove the inherent pressure pulsation component of the hydraulic pump and then invert to obtain the actual kinematic viscosity value of the current hydraulic oil, and immediately close the bypass solenoid valve after the identification is completed. S3. Compare the actual kinematic viscosity value with the reference viscosity value obtained and stored by the same identification process at the standard working oil temperature during the equipment's factory commissioning stage. Query the pre-stored relationship between viscosity and valve orifice flow gain to obtain the internal leakage compensation amount and control current bias amount of the electro-hydraulic proportional pressure valve at the current viscosity. S4. The blank holder force setting value and the main punch pressure setting value of each stretching station are superimposed with the corresponding control current bias value according to the pressure range of each station. At the same time, the current feedforward amount in the pressure building stage and the current maintenance amount in the pressure holding stage are corrected according to the internal leakage compensation amount. Combined with the pre-pressure building trigger logic, the corrected pressure setting value for the whole stroke is generated. S5. During the stamping stroke, the actual pressure value of the die holder at each station is collected by the pressure sensor at each station. The current stamping stroke stage is determined based on the slider crank angle signal. The actual pressure value of the die holder is compared with the corrected pressure set value of the corresponding station. The adjustment control quantity is calculated by combining the stage gain and control dead zone rules. The adjustment control quantity is output to distinguish between single station off-center load adjustment and overall pressure adjustment. S6. The corrected pressure setpoint and adjustment control quantity are combined with the valve characteristics under the current viscosity and converted into a current domain control signal. The hydraulic cushion actuators of each station are driven in the following manner: S-shaped slow pressure building in the pressure building stage, synchronous pressure of the main cylinder and hydraulic cushion in the pressure holding stage, and graded pressure relief in the pressure relief stage. For the rigid impact process station, only a fixed stroke ejection action is performed without pressure adjustment. After the pressure adaptive adjustment of the current stamping cycle is completed, it returns to S1.

[0006] Compared with the prior art, the beneficial effects of the present invention are as follows: By utilizing the non-stamping idle window of the slider return stroke to identify the actual kinematic viscosity of the hydraulic oil online, and combining it with the pre-stored full-temperature range valve control characteristic mapping relationship, the internal leakage compensation and control current bias are obtained. The blank holder force and main punching pressure of each station are corrected in a segmented and staged manner. With the staged adaptive gain closed-loop fine adjustment and multi-station independent hydraulic cushion drive, the problem of the control characteristics of the electro-hydraulic proportional pressure valve deviating due to changes in hydraulic oil viscosity, which in turn causes the punching pressure control accuracy to fluctuate with the working conditions, can be solved without the need to install an additional independent viscosity sensor and without occupying effective production time. It can maintain a stable punching pressure control state within different working temperatures, different oil aging degrees, and normal wear range of valve components, effectively reducing various forming quality problems caused by pressure fluctuations during the stretching process and improving the batch quality consistency of high-precision complex stretched parts. Attached Figure Description

[0007] Figure 1 This is an overall block diagram of the present invention; Figure 2 This is the overall flowchart of the present invention. Detailed Implementation

[0008] The technical solutions of this invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described invention is only a part of this invention, not all of it. All other inventions obtained by those skilled in the art based on this invention without inventive effort are within the scope of protection of this invention.

[0009] This invention provides an adaptive control system for the punching pressure of a multi-station stretching punch press. Please refer to [link / reference]. Figure 1 As shown, it includes: The stamping condition timing identification unit is used to collect the slider-crank angle signal in real time, identify the non-stamping idle window during the slider return phase based on the slider-crank angle signal, and output the identification trigger signal. The online identification unit for actual hydraulic viscosity receives the identification trigger signal and controls the opening of the master cylinder unloading valve to cut off the pressure path of the main actuator. It controls the bypass solenoid valve to open the bypass branch where the fixed throttle is located, inputs a stepped identification current sequence to the electro-hydraulic proportional pressure valve, collects multiple sets of pressure difference values ​​at different steady-state flow rates across the fixed throttle, removes the inherent pressure pulsation component of the hydraulic pump, and then inverts to obtain the actual kinematic viscosity value of the current hydraulic oil. After the identification is completed, the bypass solenoid valve is immediately closed. The valve control characteristic offset mapping unit is used to compare the actual kinematic viscosity value with the reference viscosity value obtained and stored by the same identification process at the standard working oil temperature during the equipment's factory commissioning stage. It queries the pre-stored correspondence between viscosity and valve orifice flow gain to obtain the internal leakage compensation amount and control current bias amount of the electro-hydraulic proportional pressure valve at the current viscosity. The pressure setpoint feedforward correction unit is used to superimpose the blank holder force setpoint of each stretching station with the main punch pressure setpoint and the corresponding control current bias. At the same time, it corrects the current feedforward during the pressure build-up stage and the current maintenance during the pressure holding stage based on the internal leakage compensation, and generates the corrected pressure setpoint. The stamping pressure closed-loop fine-tuning unit is used to collect the actual pressure value of the die holder during the stamping stroke, compare it with the corrected pressure set value, and output the adjustment control quantity to drive the hydraulic cushion actuator of each station to complete the adaptive adjustment of the stamping pressure.

[0010] The incremental encoder installed at the end of the crankshaft continuously outputs pulse signals as the crankshaft of the multi-station stretching punch press rotates synchronously. The controller obtains the slider crank angle signal in real time by counting the pulse signals. The incremental encoder installed at the end of the crankshaft is selected to output 10,000 pulses per revolution to ensure that the angle resolution reaches 0.03 degrees, which is sufficient to meet the accuracy requirements for slider position judgment.

[0011] When each pulse interruption is triggered, the controller reads the current slider-crank angle signal value and compares it with the pre-stored slider bottom dead center position angle value. The slider bottom dead center position angle value is the crankshaft angle value corresponding to the slider running to the lowest position, which is obtained by dial gauge calibration during the factory debugging of the equipment. When the slider-crank angle signal value is greater than the slider bottom dead center position angle value for three consecutive sampling cycles, it is determined that the slider has passed the bottom dead center and started the return motion.

[0012] The controller continuously records the change in the slider-crank angle signal value. When the slider-crank angle signal value increases to the pre-stored top dead center position angle value of the slider, it is determined that the slider has completed one return motion. The top dead center position angle value of the slider is the crankshaft angle value corresponding to the slider running to the highest position. The angle range between the bottom dead center position angle value of the slider and the top dead center position angle value of the slider is the slider return range. Within the slider return range, the controller continuously executes the window judgment logic and will not mistakenly trigger other actions outside the stamping period.

[0013] After the slider enters the slider return range, the controller begins to collect the voltage signal output by the main cylinder pressure sensor installed in the upper chamber of the main cylinder. The main cylinder pressure sensor installed in the upper chamber of the main cylinder has a range of 0 to 25 MPa, corresponding to a voltage signal of 0 to 10 volts. The controller reads the voltage signal value through a 12-bit analog-to-digital converter chip and converts it into the residual pressure value of the hydraulic system according to a linear ratio. During the conversion process, a zero-point offset correction is added to offset the fixed error caused by sensor temperature drift. The calculated residual pressure value of the hydraulic system is processed by first-order hysteresis filtering to eliminate the numerical jump caused by pipeline pressure pulsation.

[0014] While collecting the residual pressure value of the hydraulic system, the controller receives the status word sent by the multi-station feeding robot controller through the industrial bus. After completing the workpiece gripping, conveying and releasing actions of all stations, the multi-station feeding robot controller will set the corresponding bit in the status word to indicate that the workpiece transfer is complete. The workpiece transfer completion signal is the state when the corresponding bit in the status word is high.

[0015] The controller continuously checks whether three trigger conditions are met simultaneously within the slider return range. The first trigger condition is that the position corresponding to the slider crank angle signal crosses the preset angle before the top dead center position of the slider. The preset angle before the top dead center position of the slider is set to 20 degrees of crankshaft rotation angle before the top dead center, allowing enough time to complete the viscosity identification action. The second trigger condition is that the residual pressure of the filtered hydraulic system is lower than the preset pressure threshold. The preset pressure threshold is set to 0.5 MPa, indicating that the main cylinder has been completely depressurized and will not interfere with the pressure measurement of the bypass branch. The third trigger condition is that the controller receives a workpiece transfer completion signal from the multi-station feeding robot controller, indicating that the workpiece has been accurately placed in the corresponding station and the identification process will not be interrupted by the feeding action.

[0016] When all three trigger conditions are met simultaneously, the controller determines that the non-pressing idle window has entered the slider return phase and outputs an identification trigger signal to the online identification unit for actual oil viscosity. This triggers the online identification unit to begin executing the viscosity identification process. At this point, the remaining window time is calculated to ensure that the identification process is not interrupted by the next pressing action. The remaining window time calculation is based on the slider crank angle signal. First, the remaining angle between the slider and the next pressing contact position is calculated based on the current slider crank angle signal value. The next pressing contact position is the crankshaft angle value corresponding to the first contact of the punch with the sheet metal surface during the slider's descent. This crankshaft angle value is also obtained through trial pressing calibration during factory debugging. The remaining angle is divided by the real-time crankshaft angular velocity to obtain the remaining time between the slider and the next pressing contact position. The real-time crankshaft angular velocity is calculated by dividing the difference between the slider crank angle signals of two adjacent sampling periods by the sampling period. The calculation formula is as follows: , in the formula This indicates the remaining time until the slider contacts the workpiece position again during the next stamping operation. This indicates the crankshaft angle corresponding to the position where the slider will contact the workpiece during the next stamping operation. This indicates the angle value corresponding to the current slider crank angle signal. This indicates the real-time angular velocity of the crankshaft.

[0017] After calculating the remaining time before the slider contacts the workpiece position for the next stamping, it is compared with the pre-stored minimum time required for viscosity identification. The minimum time required for viscosity identification is the fixed time required for the online identification unit to complete one complete identification process for the actual viscosity of the oil. The minimum time required for viscosity identification is calibrated at the factory based on the total number of steps in the stepped identification current sequence and the steady-state holding time of each step. In this invention, the stepped identification current sequence is set with five current steps. The steady-state holding time of each step is 20 milliseconds, plus the bypass solenoid valve action time of 10 milliseconds, for a total duration of 110 milliseconds. Therefore, the minimum time required for viscosity identification is set to 120 milliseconds with a 10 millisecond margin.

[0018] If the calculated remaining time between the slider and the next press contact with the workpiece is greater than the minimum time required for viscosity identification, the non-pressing idle window during the slider return phase remains effective, allowing the online viscosity identification unit to continue the identification process. If the calculated remaining time between the slider and the next press contact with the workpiece is less than the minimum time required for viscosity identification, the non-pressing idle window during the slider return phase is immediately terminated, the identification trigger signal is stopped, and the bypass solenoid valve is closed to cut off the bypass branch where the fixed throttle is located, preventing high-pressure oil from entering the bypass branch and causing pressure shock after the next press begins.

[0019] The calculation of the remaining window time also incorporates compensation for changes in angular velocity. When a continuous increase in the real-time angular velocity of the crankshaft is detected, indicating a potential increase in the press speed, the remaining time calculation value is corrected according to the maximum acceleration to prevent the identification process from being abruptly interrupted due to the slide accelerating downwards. The correction calculation formula is as follows: in the formula This indicates the remaining time between the corrected slider position and the next press contact with the workpiece. This represents the maximum angular acceleration of the crankshaft. The value of the maximum angular acceleration of the crankshaft is calculated based on the maximum output torque of the servo motor and the moment of inertia of the flywheel and is pre-stored in the controller.

[0020] The corrected remaining time is also compared with the shortest time required for viscosity identification. As long as the corrected remaining time is less than the shortest time required for viscosity identification, the non-stamping idle window of the slider return phase ends early. This dual time judgment mechanism with angular acceleration compensation can ensure that the viscosity identification process is always in a completely undisturbed non-stamping period and will not be interrupted by the stamping action due to the fluctuation of the press speed or other abnormalities. It can effectively avoid the pressure difference measurement error caused by the impact of high pressure oil on the fixed throttle and also avoid the equipment safety risk caused by sudden stamping during the identification process.

[0021] During the non-stamping idle window of the slider return phase, the controller performs a remaining time calculation and judgment every millisecond until the online identification unit of the actual hydraulic viscosity completes a complete identification process and outputs the current actual kinematic viscosity value of the hydraulic oil. Then, the controller actively closes the non-stamping idle window of the slider return phase, or closes the non-stamping idle window of the slider return phase in advance because the remaining time is insufficient. After closing the window, the controller resets the identification trigger signal and waits for the next slider return interval to arrive before re-executing the window judgment process.

[0022] When the slider running speed changes, the window judgment logic will automatically adjust the trigger timing and window length. When the speed of the punch press increases during continuous operation, the actual time corresponding to the preset angle before the top dead center position of the slider will be shortened. The remaining time calculation will automatically identify the speed change and trigger the identification in advance to ensure sufficient identification time. When the speed of the punch press decreases during a single stroke, the non-stamping idle window during the slider return phase will be automatically extended, allowing the online identification unit of the actual viscosity of the oil to increase the average number of times and improve the measurement accuracy.

[0023] When a faulty master cylinder pressure sensor causes an abnormal residual pressure value in the hydraulic system, the trigger condition judgment will automatically prevent the window from triggering to avoid incorrect identification results from entering the subsequent control process. When the multi-station feeding robot controller is stuck and fails to send the workpiece transmission completion signal, the window will not trigger to ensure that the workpiece is completely stationary during the identification process and that the vibration caused by the workpiece movement will not affect the accuracy of the differential pressure measurement.

[0024] This window recognition method, based on multi-condition triggering, real-time remaining time calculation, and angular acceleration compensation, is more adaptable to different stamping speeds compared to the traditional fixed-angle window triggering method. At low stamping speeds, it provides a longer effective identification time to improve viscosity identification accuracy. At high stamping speeds, it automatically shortens the window length to ensure the identification action does not interfere with the normal stamping process. Simultaneous judgment of multiple triggering conditions avoids measurement errors caused by false triggering when the master cylinder is not fully depressurized or the workpiece has not been completely transferred. This increases the effective triggering accuracy of viscosity identification to over 99%, preventing false triggering or interruption of the identification process. It provides a stable and reliable time window for the subsequent online identification unit of the actual oil viscosity, ensuring accurate and reliable measurement results.

[0025] After receiving the identification trigger signal, the online identification unit of the actual oil viscosity first completes the unloading confirmation of the main cylinder. After the main cylinder unloading valve is opened, the upper chamber of the main cylinder is connected to the oil tank and the pressure of the main cylinder continues to decrease. The controller repeatedly reads the residual pressure value of the hydraulic system collected by the main cylinder pressure sensor until the residual pressure of the hydraulic system is lower than 0.2 MPa. Only after confirming that the main cylinder is completely depressurized will the subsequent current output action begin. This avoids the residual pressure of the main cylinder from entering the bypass branch and causing measurement error of the pressure difference value at both ends of the fixed throttle.

[0026] After the main cylinder pressure relief is confirmed, the controller retrieves the dead zone current value of the electro-hydraulic proportional pressure valve from the parameter area pre-stored in the valve control characteristic offset mapping unit. The dead zone current of the electro-hydraulic proportional pressure valve is the minimum operating current calibrated and stored at the factory for the electro-hydraulic proportional pressure valve. The dead zone current value of the electro-hydraulic proportional pressure valve is the minimum current value at which the proportional electromagnet starts to generate thrust to overcome the static friction of the valve core and make the valve core start to move. The rated current of the electro-hydraulic proportional pressure valve selected in this invention is 800 mA, and the dead zone current of the electro-hydraulic proportional pressure valve is calibrated to 120 mA. The dead zone current of the electro-hydraulic proportional pressure valves of different batches has dispersion. Therefore, the dead zone current value of the electro-hydraulic proportional pressure valve is individually calibrated during the commissioning of each device and stored in the non-volatile storage area of ​​the valve control characteristic offset mapping unit. It will not be lost due to power failure of the device.

[0027] After retrieving the dead zone current value of the electro-hydraulic proportional pressure valve, the dead zone current of the electro-hydraulic proportional pressure valve is used as the starting current value of the stepped identification current sequence and output to the drive board of the electro-hydraulic proportional pressure valve. The drive board converts the current value into a corresponding pulse width modulation signal and outputs it to the proportional electromagnet. The current output adopts a gradual increase method, linearly rising from 0 mA to the dead zone current value of the electro-hydraulic proportional pressure valve after five milliseconds to avoid hydraulic shock caused by sudden current changes.

[0028] After the first-stage current output, timing begins to enter the first-stage current plateau holding stage. Each current plateau holds for a pre-stored steady-state holding time. The steady-state holding time is a fixed time preset according to the hydraulic system pressure response time constant to wait for the pressure difference across the fixed throttle to reach a steady state. The hydraulic system pressure response time constant is obtained through step response test. In this invention, the hydraulic system pressure response time constant is eight milliseconds, so the steady-state holding time is set to twenty milliseconds, which is about two and a half times the time constant to ensure that the pressure difference across the fixed throttle can enter the steady-state error band.

[0029] A pressure stability determination window is set at the end of each current platform. The pressure stability determination window is a fixed time window of the last five milliseconds of each current platform. The pressure difference value across the fixed throttle is collected only within the pressure stability determination window. The fifteen milliseconds before the pressure stability determination window opens is the pressure dynamic establishment phase. The pressure difference value collected during the pressure dynamic establishment phase includes hydraulic shock fluctuations caused by current step and is not included in the subsequent viscosity calculation. Fifty pressure difference value samples across the fixed throttle are continuously collected at a sampling frequency of 10,000 Hz within the pressure stability determination window. During the collection process, gross error removal is performed on the pressure difference value samples. The average value and standard deviation of the fifty pressure difference value samples are calculated. When a pressure difference value sample deviates from the average value by more than three times the standard deviation, it is judged as a gross error sample and is removed and not included in the subsequent calculation. The average value of the remaining samples is taken as the steady-state pressure difference value corresponding to the current level.

[0030] After the first-stage current plateau holding time is reached, the output current value is increased according to the pre-stored fixed current step size, which is set to 80 mA. The current rise process of each stage also adopts a two-millisecond linear slow rise to avoid pressure shock caused by current step. After the current rises to the target value, it enters the next stage current plateau holding stage and repeats the above steady-state waiting and pressure stability judgment window sampling process.

[0031] The current level of the stepped current sequence is dynamically calculated based on the total duration. The highest current value of the stepped current sequence corresponds to the median range of the rated operating current of the electro-hydraulic proportional pressure valve. The median range of the rated operating current of the electro-hydraulic proportional pressure valve is 40% to 60% of the rated current. In this invention, the highest current value is set to 400 mA, which corresponds to 50% of the rated current. This covers the most commonly used operating pressure range of the hydraulic system and avoids additional oil temperature rise caused by high current.

[0032] Before each additional current step, the total execution time of the current step is calculated. The total execution time is the number of completed steps multiplied by the steady-state holding time of each step plus the current ramp-up time. The total execution time plus the estimated time required for the remaining steps is compared with the minimum time required for viscosity identification. When the remaining time is insufficient to complete the sampling of the next current platform, the number of current steps is stopped, and the number of completed steps is used as the effective number for subsequent calculations. This ensures that the total execution time of the entire step-type identification current sequence does not exceed the minimum time required for viscosity identification and that normal stamping operations are not affected by the timeout of the current sequence execution.

[0033] The calculation of the total step duration includes a current settling time compensation. This compensation is dynamically adjusted based on oil temperature changes. When the oil temperature is low, the hydraulic oil viscosity is high, resulting in a longer current settling time; therefore, the compensation is increased appropriately. Conversely, when the oil temperature is high, the hydraulic oil viscosity is low, resulting in a shorter current settling time; therefore, the compensation is decreased appropriately. The formula for calculating the current settling time compensation is as follows: in the formula This indicates the amount of current settling time compensation. Indicates the current settling time at the reference oil temperature. This indicates the actual kinematic viscosity value of the hydraulic oil obtained in the last identification. This indicates the reference viscosity value.

[0034] The total duration calculation after adding current settling time compensation can more accurately predict the execution time of the ladder sequence, and the actual execution time will not exceed the window range due to oil temperature changes.

[0035] After collecting the differential pressure values ​​of all effective steps, the shutdown process begins. First, the output current of the electro-hydraulic proportional pressure valve is linearly reduced from its current value to zero mA over five milliseconds. After the current returns to zero, wait for ten milliseconds to confirm that the bypass branch pressure has been completely released. Then, a control signal is output to the bypass solenoid valve to drive it to close and cut off the bypass branch where the fixed throttle is located. After the bypass solenoid valve is completely closed, a signal is output to allow the master cylinder to start building pressure for the next pressurization action. This avoids high-pressure oil impacting the fixed throttle when the master cylinder builds pressure before the bypass solenoid valve is completely closed.

[0036] The linear slow descent process, rather than direct shut-off, avoids the hydraulic shock caused by the rapid reset of the valve core due to the sudden disappearance of current. It also avoids the backflow of oil and the mixing of air bubbles caused by the sudden loss of pressure at both ends of the fixed throttle, which would affect the accuracy of the next identification.

[0037] The stepped identification current sequence will automatically adjust the number of stages under different stamping speeds. When stamping at high speed, the minimum time required for viscosity identification is shortened accordingly. At this time, the fixed current step size will be appropriately increased and the number of current stages will be reduced to ensure that at least three levels of effective current differential pressure values ​​are collected within a limited time. When stamping at low speed, the minimum time required for viscosity identification is extended accordingly. At this time, the fixed current step size will be appropriately reduced and the number of current stages will be increased to improve the viscosity inversion accuracy.

[0038] The steady-state differential pressure value collected at each current platform stage is compensated for temperature drift. Temperature changes in the fixed throttle body cause slight variations in the orifice size, affecting the accuracy of the differential pressure measurement. A thermistor installed on the fixed throttle body collects the fixed throttle body temperature in real time and corrects the differential pressure value based on the metal's thermal expansion coefficient. The correction formula is as follows: in the formula This represents the steady-state pressure difference after temperature correction. This represents the original steady-state pressure difference value collected within the pressure stability determination window. This represents the coefficient of linear expansion of the material used in a fixed throttle valve. This indicates the real-time temperature of the fixed throttle. This indicates the reference temperature when calibrating a fixed throttle.

[0039] The temperature-corrected steady-state pressure difference value is stored in a temporary buffer for use in subsequent viscosity inversion steps, preventing viscosity calculation errors caused by changes in the fixed throttle temperature. During the gross error removal process, if the number of valid samples for a certain current stage is less than thirty, the pressure difference value for that stage is deemed invalid and will not participate in subsequent viscosity calculations, ensuring that sample loss due to occasional pressure fluctuations will not affect the overall calculation accuracy.

[0040] After the bypass solenoid valve is closed, the controller will detect the residual pressure in the bypass branch. If the residual pressure in the bypass branch does not drop below 0.1 MPa within 100 milliseconds, it is determined that the bypass solenoid valve has a jamming fault and an alarm signal is output to remind the operator to repair it, so as to avoid the high pressure oil from flowing through the fixed throttle for a long time due to the fault, which will affect the measurement accuracy.

[0041] This complete current sequence control method, which starts from the dead zone current of the electro-hydraulic proportional pressure valve, linearly and gradually increases the stepped current platform, maintains pressure stability, and uses a sampling window for dynamic stage adjustment and linear gradual descent shut-off. Compared with the traditional fixed-step, fixed-stage current sequence control method, it can adapt to different speeds and oil temperatures. It ensures the measurement accuracy of the pressure difference value across the fixed throttle valve while strictly controlling the total identification time to avoid occupying the normal stamping time. The linearly increasing and decreasing current control method can eliminate hydraulic shocks and pressure fluctuations caused by sudden current changes. The pressure stability judgment window sampling method can completely avoid the dynamic pressure fluctuations during the current step process, ensuring that all collected pressure difference values ​​are steady-state values. Dynamic stage adjustment and current settling time compensation can obtain a sufficient number of effective pressure difference value samples under different operating conditions to improve the viscosity inversion accuracy. Fixed throttle valve temperature drift compensation can eliminate measurement errors caused by temperature rise after long-term operation of the fixed throttle valve. Bypass solenoid valve jamming detection can detect faults in time to avoid equipment damage. It can control the measurement repeatability error of the pressure difference value across the fixed throttle valve to within 0.5%, providing accurate raw data for subsequent viscosity inversion.

[0042] After collecting the differential pressure values ​​across all the fixed throttles at each step and compensating for the temperature of the fixed throttles, the controller extracts the differential pressure values ​​across all the fixed throttles collected within the pressure stability determination window from the temporary buffer area to form a differential pressure time series. The differential pressure time series is arranged in chronological order of sampling time, with the differential pressure value of each sampling point and its corresponding sampling timestamp.

[0043] After the differential pressure value time series is extracted, the differential pressure value series is truncated to an integer period according to the pre-stored inherent pulsation period of the hydraulic pump. The inherent pulsation period of the hydraulic pump is a fixed period value calibrated and stored at the factory based on its rated speed and the number of plungers. The axial plunger pump used in this invention has nine plungers with a rated speed of 1,500 revolutions per minute. Each time a plunger passes through the oil outlet, a pressure pulsation is generated. Therefore, the inherent pulsation period of the hydraulic pump is calculated to be 40 milliseconds. The inherent pulsation period value of the hydraulic pump is corrected by the actual pressure fluctuation frequency collected during the equipment debugging and stored in the non-volatile storage area, which will not change due to equipment restart.

[0044] During the whole cycle interception process, the pulsation phase in the pressure difference value time series is first identified. The phase difference between the pressure difference value time series and the standard pulsation waveform is found through cross-correlation calculation. The pressure difference value time series is cut into several complete hydraulic pump inherent pulsation cycle segments with the pulsation trough position as the starting point of the whole cycle interception. During the cutting process, the first and last segments that are less than a complete hydraulic pump inherent pulsation cycle are discarded and do not participate in subsequent calculations, ensuring that each segment participating in the calculation contains an integer number of pressure pulsation cycles.

[0045] After completing the full cycle, the pressure difference value within each complete hydraulic pump inherent pulsation cycle is arithmetically averaged. The periodic hydraulic pump inherent pressure pulsation component is eliminated due to positive and negative cancellation. During the arithmetic averaging process, the pulsating pressure within a complete cycle will fluctuate around the average value, and the integral result is zero. Therefore, the fixed frequency hydraulic pump inherent pressure pulsation component can be completely filtered out while retaining the true steady-state pressure change information. The average value of all complete cycles under each current level is then averaged to obtain the steady-state pressure difference value of each current level in the corresponding step-type identification current sequence.

[0046] After obtaining the steady-state pressure difference value for each current level, an abnormal data point removal operation is performed. First, the average value and standard deviation of all steady-state pressure difference values ​​under the same current level are calculated. The deviation of each steady-state pressure difference value from the average value is compared one by one. When a steady-state pressure difference value deviates from the average steady-state pressure difference value of the same level by more than a preset threshold, it is judged as an abnormal data point and removed. The preset threshold is set to three times the standard deviation, which can cover 99.7% of the normal data distribution range. Abnormal data points are usually caused by instantaneous pressure fluctuations caused by tiny impurities in the oil passing through the fixed throttle, or pressure jumps caused by the compression and expansion of tiny bubbles in the oil. If the number of remaining valid data points is less than three after the abnormal data points are removed, the corresponding current level data is determined to be invalid and will not be included in the subsequent viscosity calculation.

[0047] After the abnormal data is removed, the calibration differential pressure values ​​corresponding to each current level obtained by the same step-by-step identification current sequence calibration at the standard working oil temperature during the factory commissioning stage are retrieved from the calibration parameter area stored in the valve control characteristic offset mapping unit. The calibration differential pressure value calibration process is exactly the same as the online identification process. During the calibration process, the hydraulic oil temperature is strictly controlled at the standard working oil temperature of 40 degrees Celsius. The hydraulic oil used is the anti-wear hydraulic oil of the grade specified by the equipment. The calibration process is repeated 100 times and the average value is taken as the final calibration differential pressure value and stored in the valve control characteristic offset mapping unit.

[0048] After retrieving the calibrated differential pressure value, the steady-state differential pressure value corresponding to each current level is compared point by point with the calibrated differential pressure value under the same current. The differential pressure deviation for each current level is obtained by subtracting the calibrated differential pressure value from the steady-state differential pressure value under the same current level. A positive differential pressure deviation indicates that the current differential pressure value is greater than the calibrated differential pressure value, which means that the current hydraulic oil viscosity is greater than the reference viscosity value. A negative differential pressure deviation indicates that the current differential pressure value is less than the calibrated differential pressure value, which means that the current hydraulic oil viscosity is less than the reference viscosity value.

[0049] After obtaining the pressure difference deviation for each current level, the viscosity deviation for each level is calculated based on the pre-stored relationship between the flow rate, pressure difference, and viscosity of the fixed throttle. This relationship is derived from the laminar flow throttling formula. Under laminar flow conditions, the pressure difference across the fixed throttle is directly proportional to the hydraulic oil viscosity. The formula for calculating the viscosity deviation, considering the Reynolds number correction at the fixed throttle inlet fillet, is as follows: in the formula Indicates the first The viscosity deviation value corresponding to the current level. Indicates the reference viscosity value. Indicates the first The steady-state voltage difference corresponding to the stage current, Indicates the first The calibrated differential voltage value corresponding to the current level. This represents the correction factor for the inlet fillet radius of the fixed throttle, a dimensionless constant obtained from the calibration of the inlet fillet radius of the fixed throttle. Indicates the first The Reynolds number of the oil in the fixed throttle under a fixed current.

[0050] After calculating the viscosity deviation values ​​corresponding to each current level, the viscosity deviation values ​​of each level are weighted according to the pre-stored weighted average coefficient. The weighted average coefficient is the pre-stored weight value corresponding to different current ranges. The principle of setting the weighted average coefficient is that the weighted average coefficient of the flow range in the electro-hydraulic proportional pressure valve is higher than that of the small flow range and the large flow range. In the small flow range, the oil flow in the fixed throttle is affected by the laminar flow initiation section, and the measurement error is large. Therefore, the weight is set to 0.1. In the medium flow range, the oil flow is fully developed laminar flow, and the measurement accuracy is the highest. Therefore, the weight is set to 0.7. In the large flow range, the oil flow in the fixed throttle begins to transition to turbulent flow, and the measurement error increases. Therefore, the weight is set to 0.2. The sum of all weighted average coefficients is one to ensure that the weighted average result is unbiased.

[0051] The weighted average process also includes a consistency check of adjacent levels of data. If the viscosity deviation value calculated at a certain level deviates from the average viscosity deviation value of the two adjacent levels by more than 20%, it is determined that there is a local anomaly in the corresponding level of data, and the weight of the corresponding level is reset to zero and it is not included in the weighted calculation, so as to avoid the overall calculation deviation caused by the anomaly of a single level of data.

[0052] After obtaining the overall viscosity deviation value through weighted averaging, the actual kinematic viscosity value of the current hydraulic oil is obtained by adding the overall viscosity deviation value to the reference viscosity value. The calculation formula is as follows: in the formula This indicates the current actual kinematic viscosity value of the hydraulic oil. Indicates the first The weighted average coefficient corresponding to the current level. Indicates the effective current level.

[0053] After calculating the actual kinematic viscosity of the current hydraulic oil, a rationality check is first performed. The actual kinematic viscosity of the current hydraulic oil is compared with the actual kinematic viscosity value obtained in the previous identification. If the difference between the two viscosity values ​​exceeds 20%, the identification result is deemed abnormal and discarded. The actual kinematic viscosity value obtained in the previous identification is used and output to the valve control characteristic offset mapping unit. If the difference between the two viscosity values ​​is within 20%, the actual kinematic viscosity value of the current hydraulic oil is stored in the non-volatile storage area to update the historical viscosity value. At the same time, it is output to the valve control characteristic offset mapping unit to query the corresponding internal leakage compensation amount and control current bias amount. The rationality check can avoid the erroneous identification results caused by accidental factors from entering the control loop and causing pressure control abnormalities.

[0054] The actual kinematic viscosity value of the hydraulic oil is output and then used to update the current settling time compensation calculation for the next identification, forming a calculation closed loop to improve the accuracy of subsequent time compensation. When the hydraulic oil is cold-started, the temperature is low and the viscosity is high. At this time, the calculated actual kinematic viscosity value will be significantly greater than the reference viscosity value. The valve control characteristic offset mapping unit will output a larger control current bias based on the larger viscosity value to compensate for the valve flow loss caused by the high viscosity of the oil at low temperature. As the equipment runs continuously, the oil temperature rises and the viscosity decreases, the actual kinematic viscosity value will gradually decrease, and the control current bias will also decrease accordingly to avoid overcompensation.

[0055] Changes in viscosity caused by the aging of hydraulic oil over long-term use can also be accurately identified using the above calculation method without the need for manual recalibration of parameters. Abnormal viscosity changes caused by mixing different grades of hydraulic oil can also be accurately identified using the above calculation method without the need for manual recalibration of parameters.

[0056] During the factory commissioning phase of the equipment, the first step is to calibrate the relationship between viscosity and valve flow gain. The calibration process is completed on a dedicated hydraulic test bench. The hydraulic station of the multi-station stretching punch press is placed in a high and low temperature environment chamber. Multiple standard working oil temperature points are set in the full temperature range covering cold start-up to continuous operation. The standard working oil temperature points are set to five temperature points: 20 degrees Celsius, 30 degrees Celsius, 40 degrees Celsius, 50 degrees Celsius, and 60 degrees Celsius, covering the entire working range from the cold start-up oil temperature to the highest oil temperature during continuous operation.

[0057] After the ambient temperature reaches each standard working oil temperature point, it is kept at that temperature for two hours to ensure that the hydraulic oil temperature is uniform and stable, with temperature fluctuations not exceeding 0.5 degrees Celsius. At each standard working oil temperature point, the same step-by-step identification current sequence and differential pressure value processing procedure as the online identification is adopted. Starting from the dead zone current of the electro-hydraulic proportional pressure valve, the current is gradually increased to the median of the rated current. Each current level is maintained for a sufficient time to wait for the pressure to stabilize. The differential pressure value across the fixed throttle is collected, and after filtering out the inherent pulsation component of the hydraulic pump and removing outliers, the kinematic viscosity value of the hydraulic oil at the corresponding temperature is calculated as the calibration viscosity. The calibration viscosity is the kinematic viscosity value of the hydraulic oil measured at each standard working oil temperature point through the same identification procedure. The viscosity measured at the 40-degree Celsius standard working oil temperature reference point is the reference viscosity value, which corresponds to the nominal kinematic viscosity marked on the hydraulic oil grade.

[0058] The output pressure value of the electro-hydraulic proportional pressure valve under different control currents was obtained by repeating the test fifty times at each calibrated viscosity point. The relationship between current and pressure was fitted into a smooth curve to obtain a complete family of valve control characteristic curves of the electro-hydraulic proportional pressure valve at each calibrated viscosity. The family of valve control characteristic curves is stored in the non-volatile storage area of ​​the valve control characteristic offset mapping unit.

[0059] After calibration, characteristic parameters for each calibrated viscosity are extracted from the valve control characteristic curve family and pre-stored in the viscosity-valve flow gain correspondence. The extracted characteristic parameters include zero-point leakage current value, valve flow gain coefficient for the entire pressure range, pressure build-up feedforward coefficient, and pressure holding coefficient. The zero-point leakage current value is the minimum control current value corresponding to the output pressure of the electro-hydraulic proportional pressure valve rising from zero. The zero-point leakage current value decreases as viscosity decreases because the internal leakage at the valve orifice increases after viscosity decreases, requiring a smaller current to start pressure build-up. The valve flow gain coefficient is the rate of change of output pressure of the electro-hydraulic proportional pressure valve corresponding to a unit change in control current. The entire pressure range is divided into pressure ranges of two MPa according to pressure magnitude, and the valve flow gain coefficient for each range is calculated separately. The pressure build-up feedforward coefficient is the proportional coefficient of the additional current required during the pressure build-up stage, and the pressure holding coefficient is the proportional coefficient of the additional current required during the pressure holding stage. Both the pressure build-up feedforward coefficient and the pressure holding coefficient increase as viscosity decreases because the internal leakage increases with viscosity, requiring more additional current to compensate for the leakage.

[0060] During online operation, the online viscosity identification unit calculates the actual kinematic viscosity value of the hydraulic oil and sends it to the valve control characteristic offset mapping unit. The valve control characteristic offset mapping unit uses the actual kinematic viscosity value of the hydraulic oil as an index to find two adjacent calibration viscosity points among the calibration viscosity points. One calibration viscosity point is greater than the actual kinematic viscosity value of the hydraulic oil, and the other is less. Linear interpolation is then used between the two adjacent calibration viscosity points to calculate the zero-point leakage current value at the current viscosity and the valve flow gain coefficient for each pressure range. The linear interpolation formula is as follows: in the formula This indicates the parameter values ​​that need to be interpolated at the current viscosity, including the zero-point leakage current value and the valve orifice flow gain coefficient for each pressure range. This indicates the parameter value corresponding to the low viscosity calibration point. This indicates the parameter value corresponding to the high viscosity calibration point. This indicates the current actual kinematic viscosity value of the hydraulic oil. This indicates the calibration viscosity value corresponding to adjacent low viscosity calibration points. This indicates the calibration viscosity value corresponding to adjacent high viscosity calibration points.

[0061] A viscosity change rate constraint is added to the linear interpolation process. If the change rate of the actual kinematic viscosity value obtained between two consecutive identifications exceeds two square millimeters per second, a viscosity jump is identified. A first-order hysteresis filter is applied to the interpolation result to avoid pressure fluctuations caused by abrupt parameter changes. The filter calculation formula is as follows: in the formula This represents the parameter value output after this filtering. This represents the parameter value obtained from this interpolation. This represents the parameter value output after the previous filtering. This indicates that the filter coefficient is set to 0.2.

[0062] After interpolating to obtain the zero-point leakage current value at the current viscosity and the valve flow gain coefficient for each pressure range, parameter conversion is performed to obtain the internal leakage compensation and control current bias. The zero-point leakage current value is combined with the pressure build-up feedforward coefficient and the pressure holding coefficient to convert it into the internal leakage compensation. The internal leakage compensation includes two parts: the pressure build-up compensation component and the pressure holding compensation component. The pressure build-up compensation component is the zero-point leakage current value multiplied by the pressure build-up feedforward coefficient, which is used to correct the current feedforward during the pressure build-up stage. The pressure holding compensation component is the zero-point leakage current value multiplied by the pressure holding coefficient, which is used to correct the current holding during the pressure holding stage. After the internal leakage compensation is calculated, it is sent to the pressure setpoint feedforward correction unit for feedforward correction.

[0063] The valve flow gain coefficients for each pressure range at the current viscosity are compared with the reference flow gain coefficient at the reference viscosity. The reference flow gain coefficient is the valve flow gain coefficient for the corresponding pressure range at the reference viscosity. The difference between the two is converted into a control current bias according to the pressure range. The conversion formula is as follows: in the formula Indicates the first The control current bias corresponding to each pressure range Indicates the first The pressure setpoints corresponding to each pressure range. Indicates the number of times the viscosity is at the current level. Valve orifice flow gain coefficient for each pressure range Indicates the first viscosity at the reference viscosity The reference flow gain coefficient for each pressure range.

[0064] The control current bias is calculated separately for each pressure range and sent to the pressure setpoint feedforward correction unit. This is used to superimpose the blank holder force setpoint and main punch pressure setpoint at each stretching station. Parameter limiting protection is also added during parameter conversion. The maximum internal leakage compensation does not exceed 20% of the rated current of the electro-hydraulic proportional pressure valve, and the maximum control current bias does not exceed 15% of the rated current of the electro-hydraulic proportional pressure valve, preventing sensor malfunctions from causing abnormal parameters and pressure runaway. During cold start-up, the hydraulic oil temperature is low and viscosity is high, resulting in a larger interpolated zero-point leakage current value and a higher valve orifice flow gain coefficient. The calculated internal leakage compensation and control current bias are both negative, indicating a need to reduce the control current to avoid excessive pressure. As the equipment runs continuously, the oil temperature rises and viscosity decreases, leading to a decrease in the zero-point leakage current value and a lower valve orifice flow gain coefficient. The internal leakage compensation and control current bias gradually become positive, indicating a need to increase the control current to compensate for insufficient pressure caused by internal leakage and decreased gain.

[0065] As hydraulic oil ages and its viscosity decreases over time, the control current bias gradually increases. When the control current bias continuously exceeds 10% of the rated current, a hydraulic oil replacement reminder signal is output to prompt the user to replace the hydraulic oil. Linear interpolation between different viscosity points ensures that parameters change continuously without jumps. First-order hysteresis filtering further smooths the parameter change process. Calculating the control current bias across different pressure ranges can adapt to the nonlinear changes in valve port gain under different pressures. Compared with existing technologies that use single gain compensation, this reduces the pressure control error from 10% to less than 2%. Parameter limiting protection prevents pressure runaway under abnormal conditions. Full temperature range calibration covers all operating conditions from cold start to long-term continuous operation without manual parameter adjustment. The hydraulic oil aging reminder function prompts the user to perform timely maintenance to avoid product quality problems caused by deterioration of oil performance.

[0066] After the valve control characteristic offset mapping unit outputs the internal leakage compensation amount and the control current bias amount for each pressure range, the pressure setpoint feedforward correction unit starts to execute the pressure setpoint correction process. First, the complete stamping stroke is divided into four stages according to the slider crank angle signal. The process is as follows: from the beginning of the slider's descent to the point where the punch contacts the plate surface, it is the material contact stage; from the point where the punch contacts the plate to the point where the pressure reaches the set value, it is the pressure building and stretching stage; from the point where the pressure reaches the set value and is maintained until the slider begins to return, it is the pressure holding and shaping stage; and from the point where the slider begins to return to the top dead center, it is the pressure relief and return stage. The four stages are divided according to the fixed angle range corresponding to the slider crank angle signal. The angle range is calibrated and stored during equipment debugging based on the die closing height during trial stamping.

[0067] After the stroke stages are divided, the blank holder force setting value and the main punch pressure setting value of each stretching station are mapped to the pre-stored pressure range mapping table according to the pressure magnitude. The pressure range mapping table is a pressure segment correspondence table that corresponds one-to-one with the valve port flow gain coefficient of each pressure range. The pressure range mapping table is divided into a range of two MPa from zero to the maximum working pressure. Each range corresponds to a control current bias and a set of valve port flow gain coefficients. The pressure range mapping table corresponds one-to-one with the pressure range parameters output by the valve control characteristic offset mapping unit, so there will be no range mismatch problem. The blank holder force setting for each stretching station is a fixed blank holder force value preset according to the process requirements of different stretching stations. For the blank holder force setting value of each stretching station, the pressure range in which the pressure magnitude is located is first determined. Based on the pressure range, the control current bias amount of the corresponding range is called and independently superimposed. Different stretching stations match the corresponding control current bias amount according to the range in which their blank holder force setting value is located. The blank holder force of the first deep stretching station is larger and belongs to the high pressure range, so the control current bias amount corresponding to the high pressure range is called. The blank holder force of the subsequent shaping station is smaller and belongs to the low pressure range, so the control current bias amount corresponding to the low pressure range is called. The blank holder force setting value of each stretching station is superimposed independently and does not affect each other, avoiding the problem of overcompensation of low pressure stations and undercompensation of high pressure stations caused by uniformly superimposing fixed bias amounts.

[0068] The main stroke pressure setting value is not a fixed value; it changes dynamically with the slider-crank angle signal. For the main stroke pressure setting value that changes dynamically with the slider-crank angle signal, the control current bias of the pressure range where the main stroke pressure is located at the current stroke position is matched point by point according to the fixed crank angle step size. The fixed crank angle step size is set to 0.5 degrees crankshaft angle. At each step size position, the current main stroke pressure setting value is read to determine the pressure range. The control current bias of the corresponding range is called and superimposed on the main stroke pressure setting value at the current position. After the point-by-point superposition is completed, the bias correction curve of the main stroke pressure setting value is obtained.

[0069] During the point-to-point stacking process, a smooth transition is implemented at the stage switching points. The control current bias of five steps before and after the switching point between the pressure building and stretching stages and the pressure holding and shaping stages uses a linear transition to avoid step-like jumps in the pressure setpoint. A ten-step linear transition is also used at the boundary positions of different pressure intervals. The calculation formula for the smooth transition at the interval boundaries is as follows: in the formula Indicates the transition section number The control current bias amount in one step. This indicates the control current bias amount for the interval before switching. This indicates the control current bias amount for the switched interval. Indicates the step size number within the transition section. This indicates that the total step size of the transition section is ten.

[0070] During the offset stacking process, the pressure setting value of the rigid impact process station performing edge trimming, punching, and blanking is not stacked with the control current offset. The pressure of the rigid impact process station is guaranteed by mechanical limit and does not require precise pressure control. Stacking the control current offset will cause excessive pressure and damage the mold. Therefore, the rigid impact station directly uses the original pressure setting value without correction to avoid overcompensation under rigid impact conditions.

[0071] After the offsets of the blank holder force setting and the main punch pressure setting are superimposed on all stretching stations, internal leakage compensation correction begins. The current feedforward in the pressure build-up stage and the current maintenance in the pressure holding stage are corrected according to the internal leakage compensation. The current feedforward in the pressure build-up stage is the extra pressure value injected in advance at the end of the material contact stage. The original value of the current feedforward in the pressure build-up stage is a fixed value, which is proportionally corrected according to the pressure build-up compensation component in the internal leakage compensation. The corrected current feedforward in the pressure build-up stage is superimposed on the main punch pressure setting and the corresponding blank holder force setting at the end of the material contact stage to build pressure in advance and compensate for the pressure build-up lag caused by internal leakage. The current maintenance in the pressure holding stage is the extra pressure value continuously injected in the pressure holding and shaping stage. The original value of the current maintenance in the pressure holding stage is a fixed value, which is proportionally corrected according to the pressure holding compensation component in the internal leakage compensation. The corrected current maintenance in the pressure holding stage is continuously superimposed on the main punch pressure setting and the corresponding blank holder force setting throughout the pressure holding and shaping stage to compensate for the pressure drop caused by internal leakage.

[0072] During the internal leakage compensation and correction process, a smooth transition is also carried out at the stage switching position. The feedforward injection from the material contact stage to the pressure building and stretching stage adopts a linear gradual increase of five degrees of crankshaft rotation angle. The feedforward withdrawal from the pressure building and stretching stage to the pressure holding and shaping stage adopts a linear gradual decrease of three degrees of crankshaft rotation angle. At the beginning of the pressure holding and shaping stage, the pressure holding amount adopts a linear increase of two degrees of crankshaft rotation angle to avoid pressure sudden change.

[0073] After completing the offset superposition and internal leakage compensation correction, the final corrected pressure setting value is generated. The corrected pressure setting value includes the independent blank holder force setting value curve and the main punch pressure setting value curve for each stretching station. All curves correspond one-to-one with the change of the slider crank angle signal and the punching stroke.

[0074] After the corrected pressure setpoint is generated, its rationality is verified. It is checked whether the pressure setpoint at each position is within the maximum pressure range of the hydraulic system, whether the pressure change rate between adjacent steps exceeds the maximum allowable value, and whether the pressure fluctuation range during the pressure holding stage is within the allowable range of the process. If the pressure exceeds the limit or the change rate exceeds the limit, the pressure setpoint at the corresponding position is limited to the allowable range to avoid pressure shock. If the pressure fluctuation during the pressure holding exceeds the limit, the pressure holding amount at the corresponding position is appropriately increased to ensure pressure stability.

[0075] When the machine is cold-started, the hydraulic oil viscosity is high and the control current bias is negative. After the bias is applied, the pressure setting value is slightly lower than the original setting value to avoid excessive pressure that could cause the workpiece to crack. As the equipment operates continuously, the oil temperature rises and the viscosity decreases, and the control current bias becomes positive. After the bias is applied, the pressure setting value is slightly higher than the original setting value to compensate for insufficient pressure caused by internal leakage and decreased gain. As the hydraulic oil ages and its viscosity continues to decrease, the control current bias will gradually increase to ensure stable pressure until an oil change reminder is triggered.

[0076] The method of independently superimposing offset values ​​at different workstations can adapt to the correction requirements of different pressures at multiple workstations. The point-by-point superposition of the main punch pressure curve can adapt to the dynamic changes of the main pressure. The smooth transition of stage switching and interval boundaries can eliminate the pressure steps caused by parameter switching. The absence of compensation at rigid impact workstations can avoid overcompensation and damage to the mold. The staged smooth injection of internal leakage can simultaneously solve the problems of pressure build-up lag and pressure drop during holding. The rationality verification can avoid equipment damage caused by abnormal parameters. Compared with the existing technology of correcting the overall offset setting value, the pressure control error at different workstations and stroke positions can be controlled within one percent. It effectively solves the problem of wrinkling, cracking and dimensional deviation of multi-station drawn parts caused by viscosity changes. The corrected pressure setting value is sent to the stamping pressure closed-loop fine-tuning unit as the target value for closed-loop control.

[0077] After the control current bias of the blank holder force setting value and the main punch pressure setting value of each stretching station is superimposed point by point, the pre-pressure building and pressure holding injection process for internal leakage compensation is started. First, the pre-pressure building trigger position is determined according to the slider crank angle signal. The pre-pressure building trigger position is the position of the fixed crank angle before the slider runs to the material contact stage. The fixed crank angle base value is the fixed position calibrated by the equipment at the factory according to different punching strokes. Different dies with different closing heights correspond to different base angle values. After the die is replaced, the base value of the pre-pressure building trigger position is automatically calibrated by trial punch.

[0078] After the baseline value of the pre-pressure build-up trigger position is calibrated, it is dynamically corrected based on the actual kinematic viscosity of the hydraulic oil. Higher viscosity results in greater hydraulic oil flow resistance and a slower pressure build-up response, requiring earlier triggering of the pre-pressure build-up. Conversely, lower viscosity leads to a faster pressure build-up response, allowing for a more appropriate delay in triggering the pre-pressure build-up to avoid increased energy consumption during idle strokes caused by premature pressure build-up. The formula for calculating the dynamically corrected pre-pressure build-up trigger angle is as follows: in the formula This indicates the crankshaft angle corresponding to the corrected pre-build pressure trigger position. This indicates the base angle of the pre-set pressure trigger position as specified by the manufacturer. This represents the pre-pressure build-up angle correction coefficient, which is calibrated based on the pressure build-up response characteristics of the hydraulic system. This indicates the current actual kinematic viscosity value of the hydraulic oil. This indicates the reference viscosity value.

[0079] After determining the corrected pre-pressure building trigger position, the slider crank angle signal is monitored in real time during slider operation. When the slider crank angle signal reaches the pre-pressure building trigger position, the current feedforward amount for the pressure building stage is injected. The current feedforward amount for the pressure building stage is obtained by multiplying the internal leakage compensation amount by the pressure building feedforward coefficient of the corresponding station. The pressure building feedforward coefficient is a fixed coefficient pre-stored according to the process requirements of different stretching stations. The first deep stretching station has a long stretching stroke and a large material inflow, which requires a high initial blank holder force. Therefore, the pressure building feedforward coefficient of the first deep stretching station is higher than that of the subsequent shaping station. The pressure building feedforward coefficient of the first deep stretching station is set to 1.2, the pressure building feedforward coefficient of the subsequent second stretching station is set to 1.1, and the pressure building feedforward coefficient of the final shaping station is set to 1.0. The current feedforward amount for the pressure building stage of different stations is calculated and injected independently without affecting each other.

[0080] The current feedforward injection process during the pressure build-up phase adopts a linear gradual increase method. From the pre-pressure build-up trigger position to the end of the material contact phase, the current feedforward increases linearly from zero to the calculated value within the angle range. This avoids hydraulic shock caused by sudden injection pressure. The slope of the linear gradual increase is automatically calculated based on the angle length of the pre-pressure build-up interval to ensure that the current feedforward during the pressure build-up phase reaches the target value when contacting the material.

[0081] After the slider enters the pressure building and stretching stage, the current feedforward of the pressure building stage remains constant until the pressure building is completed. After entering the pressure holding and shaping stage, the current feedforward of the pressure building stage linearly decreases to zero after three crankshaft rotations. At the same time, the current maintenance amount of the pressure holding stage begins to be continuously injected. The current maintenance amount of the pressure holding stage is obtained by multiplying the internal leakage compensation amount by the pressure holding maintenance coefficient of the corresponding station. It is used to compensate for the pressure drop caused by internal leakage during the pressure holding process. Different stretching stations pre-store the corresponding pressure holding maintenance coefficient according to the pressure holding time requirements. The pressure holding time of the first deep stretching station is long and the internal leakage accumulation is large, so the pressure holding maintenance coefficient is set to 1.3. The pressure holding time of the second stretching station is medium and the pressure holding maintenance coefficient is set to 1.15. The pressure holding time of the shaping station is short and the pressure holding maintenance coefficient is set to 1.0.

[0082] During the pressure holding phase, an adaptive correction for pressure drop is incorporated into the current injection process. This is achieved by dynamically fine-tuning the pressure holding coefficient based on the real-time detection of the pressure drop rate from the previous pressure holding phase. When the pressure drop rate exceeds a threshold, the coefficient is appropriately increased; conversely, when the pressure drop rate is negative and pressure increases, the coefficient is appropriately decreased. The adaptive correction formula is as follows: in the formula This indicates the pressure holding coefficient used in this stamping process. This indicates the pressure holding factor used in the previous stamping. Indicates the adaptive correction step size. This indicates the actual pressure drop during the previous pressing and holding phase. This indicates the target pressure reduction value.

[0083] The current maintenance during the pressure holding stage is also limited, not exceeding 10% of the corresponding station's blank holder force setting, to avoid excessive pressure due to overcompensation. The current maintenance during the pressure holding and shaping stage is continuously injected until the pressure holding ends. After entering the pressure release and return stage, the current maintenance during the pressure holding stage linearly decreases to zero after two crankshaft rotations to avoid pressure release impact. For rigid impact processes, no current feedforward or current maintenance during the pressure building stage is injected. In rigid impact processes, the punching is completed instantly after contact with the material, requiring neither pre-pressure building nor pressure holding. Injecting feedforward and maintenance currents is actually... This can cause excessive punching force and damage to the mold. The current feedforward amount during the pressure building stage and the current maintenance amount during the pressure holding stage are injected simultaneously with the previously completed control current bias amount superposition process. The control current bias amount is superimposed on the basic pressure setting value. The current feedforward amount during the pressure building stage and the current maintenance amount during the pressure holding stage are superimposed on the pressure setting value at the corresponding position according to the stroke position. After all the correction amounts are superimposed, the corrected pressure setting value for the entire stroke is generated. The corrected pressure setting value is stored in array form according to the crankshaft rotation angle step of 0.5 degrees. Each step corresponds to a main punch pressure value and the blank holder force value of each stretching station.

[0084] After generating the corrected pressure setpoint, a full-process rationality check is performed. This check includes verifying whether the pressure value at each angle position is within the allowable range of the hydraulic system, whether the pressure change rate between adjacent steps exceeds the maximum allowable value, whether the pressure rise slope during the pressure build-up stage meets the process requirements, and whether the pressure fluctuation during the pressure holding stage is within the allowable range. If any abnormal value occurs, it is automatically corrected to the allowable range.

[0085] When the machine is started cold, the hydraulic oil viscosity is high, and the pre-pressure building trigger position is automatically advanced. The current feedforward during the pressure building stage and the current maintenance during the pressure holding stage are increased accordingly to compensate for the large leakage at low temperature. After the equipment runs continuously, the oil temperature rises and the viscosity decreases, and the pre-pressure building trigger position is automatically delayed. The compensation amount is reduced accordingly to avoid overcompensation. After the equipment runs for a long time, the internal leakage of valve wear increases, and the pressure holding coefficient will gradually increase through adaptive correction to automatically compensate for the increase in leakage caused by wear. The corrected pressure setpoint is sent to the stamping pressure closed-loop fine-tuning unit as the target curve for closed-loop control.

[0086] After the corrected pressure setpoint is issued, the stamping stroke closed-loop control stage begins. During the slide's descent, the actual pressure value of the die holder at each station is collected by the station pressure sensor installed between the die holder and the worktable at a sampling frequency of 5 kHz. The collected actual pressure value of the die holder is filtered by a second-order Butterworth low-pass filter to eliminate high-frequency interference caused by hydraulic shock and mechanical vibration. The filter cutoff frequency is set to 200 Hz to retain the effective pressure control signal while filtering out noise.

[0087] The filtered actual pressure value of the die holder is thinned and aligned with the corrected pressure setting value step size according to the crankshaft angle step size of 0.5 degrees. The current stamping stroke stage is determined according to the slider crank angle signal. Different stamping stroke stages call the corresponding stage pre-stored stage gain coefficient. The pressure building and stretching stage uses low gain to avoid pressure overshoot. The pressure holding and shaping stage uses high gain to ensure pressure accuracy. The pressure release and return stage has zero gain and no closed-loop adjustment is performed.

[0088] After the basic value of the phased gain coefficient is calibrated, it is adaptively adjusted based on the current pressure deviation and the actual kinematic viscosity of the hydraulic oil. When the pressure deviation is large, the gain is increased appropriately to accelerate the response speed; when the pressure deviation is small, the gain is decreased appropriately to avoid pressure oscillation. When the viscosity is high, the valve response is slow, so the gain is increased appropriately; when the viscosity is low, the valve response is fast, so the gain is decreased appropriately. The adaptive gain calculation formula is as follows: in the formula This represents the adaptively adjusted proportional gain. This represents the base proportional gain for the corresponding stage. This represents the deviation correction factor. This indicates the current pressure deviation value. This represents the viscosity correction factor. This indicates the current actual kinematic viscosity value of the hydraulic oil. This indicates the reference viscosity value.

[0089] After the gain adjustment is completed, the actual pressure value of the die holder at each stretching station is compared with the corrected pressure setting value of the corresponding station point by point. The pressure deviation value is calculated by subtracting the actual pressure value of the die holder from the corrected pressure setting value at the same crankshaft angle position. A positive pressure deviation value indicates that the actual pressure is lower than the setting value and the control amount needs to be increased. A negative pressure deviation value indicates that the actual pressure is higher than the setting value and the control amount needs to be decreased.

[0090] After calculating the pressure deviation value, the first step is to determine whether it is within the control dead zone. The control dead zone threshold is not a fixed value but is dynamically adjusted according to the current viscosity. When the viscosity is high, the valve response has hysteresis, so the control dead zone threshold should be increased appropriately to avoid frequent adjustments. When the viscosity is low, the valve response is sensitive, so the control dead zone threshold should be decreased appropriately to improve control accuracy. The dynamic control dead zone threshold is also calculated based on the actual kinematic viscosity value. When the pressure deviation value is within the dynamic control dead zone threshold range, no adjustment control quantity is output to avoid frequent operation of the electro-hydraulic proportional pressure valve and the hydraulic cushion actuators at each station, which would cause pressure oscillations and extend the service life of the valve components.

[0091] After the pressure deviation value exceeds the control dead zone threshold, the deviation type is further determined. First, the average value of the pressure deviation values ​​of all stretching stations is calculated to obtain the overall pressure deviation. Then, the difference between the pressure deviation value of each stretching station and the overall pressure deviation is calculated to obtain the single-station off-center load deviation. When the single-station off-center load deviation of a single stretching station exceeds the pre-stored off-center load deviation threshold, it is determined to be a single-station pressure deviation caused by multi-station off-center load. The adjustment control quantity is output separately to the hydraulic cushion actuator of the corresponding station, while the control quantities of other stations remain unchanged to compensate for the single-station pressure deviation caused by multi-station off-center load. When the pressure deviation values ​​of all stations are in the same direction and the difference between them and the overall pressure deviation is less than the off-center load deviation threshold, it is determined to be an overall pressure deviation. The adjustment control quantity is output to the electro-hydraulic proportional pressure valve of the main cylinder to adjust the pressure of the main cylinder.

[0092] Adjusting the control quantity using incremental calculation avoids integral saturation. The formula for calculating the incremental control quantity is as follows: in the formula This represents the incremental adjustment control quantity output this time. The above is the adaptively adjusted proportional gain. For integral gain, For differential gain, This is the pressure deviation value for this test. This is the previous pressure deviation value. The incremental adjustment control quantity is calculated based on the previous pressure deviation value and then output limited. The maximum single adjustment control quantity shall not exceed two percent of the rated current to avoid excessive adjustment and pressure fluctuation. The output adjustment control quantity is superimposed on the feedforward corrected control quantity to drive the electro-hydraulic proportional pressure valve and the hydraulic cushion actuator of each station. For the rigid impact process station, only the actual pressure value of the mold base is collected for over-limit shutdown protection, and no adjustment control quantity is output. When the actual pressure value of the mold base of the rigid impact process station exceeds the maximum allowable pressure, an emergency stop signal is immediately output to stop the slide block operation and protect the mold and equipment. During normal stamping, no pressure fine adjustment is performed on the rigid impact process station to avoid erroneous adjustment under impact conditions.

[0093] During the closed-loop control process, integral anti-saturation processing is continuously performed. When the control output reaches the limit value, the accumulation of the integral term is stopped to avoid pressure overshoot caused by integral saturation. During the pressure holding and shaping stage, the integral gain is appropriately increased to eliminate static error and ensure pressure stability. During the pressure building and stretching stage, the derivative gain is appropriately reduced to avoid erroneous adjustment caused by derivative spikes during the pressure rise process.

[0094] When the machine is cold-started, the hydraulic oil viscosity is high. The control dead zone threshold is automatically increased, and the proportional gain is appropriately increased to compensate for the response lag caused by the hysteresis of the compensation valve. After the oil temperature rises and the viscosity decreases during continuous operation, the control dead zone threshold is automatically decreased, and the proportional gain is appropriately reduced to avoid pressure oscillation. When there is an off-center load in a single station, only the hydraulic cushion of the corresponding station is adjusted without affecting the normal pressure of other stations. When there is an overall pressure deviation, the main cylinder pressure is adjusted to ensure that the pressure of all stations is adjusted synchronously, effectively solving the quality problems of tensile parts caused by viscosity changes and off-center loads.

[0095] After the stamping pressure closed-loop fine-tuning unit outputs the adjustment control quantity, it enters the actuator drive stage. First, the corrected pressure setpoint in the pressure domain is combined with the adjustment control quantity and the characteristics of the electro-hydraulic proportional pressure valve under the current viscosity to convert it into a current domain control signal. The conversion process uses the valve orifice flow gain and internal leakage compensation parameters output by the valve control characteristic offset mapping unit under the current viscosity, and does not use the fixed conversion coefficient under the reference viscosity.

[0096] Hysteresis compensation is added during the pressure-to-current conversion process. Electro-hydraulic proportional pressure valves exhibit hysteresis, where the current rise and fall characteristics are inconsistent. The hysteresis width changes with viscosity; it is larger at higher viscosity and smaller at lower viscosity. The hysteresis compensation is calculated based on the direction of change of the current control quantity and the actual kinematic viscosity of the hydraulic oil. The hysteresis compensation is subtracted when the current rises and added when the current falls. The formula for calculating the hysteresis compensation is as follows: in the formula This represents the hysteresis compensation current. Indicates the hysteresis width at the reference viscosity. This indicates the current actual kinematic viscosity value of the hydraulic oil. Indicates the reference viscosity value. The sign indicates the direction of pressure change; positive for increasing pressure and negative for decreasing pressure.

[0097] After obtaining the current domain control signal, different drive strategies are executed according to the stamping stroke stage. During the pressure building and stretching stage, the hydraulic cushion actuators at each station are driven to build pressure according to a pre-stored S-shaped gradual rise curve. The S-shaped gradual rise curve adopts a sinusoidal acceleration and deceleration law. In the initial stage of pressure building, the pressure rise rate gradually increases; in the middle stage, the pressure rises at a constant rate; and in the final stage, the pressure rise rate gradually decreases until the set pressure is reached. The pressure change rate throughout the entire pressure building process does not exceed the pre-stored pressure change rate limit value. The pressure change rate limit value is calibrated according to the material forming process to avoid local stress concentration caused by sudden pressure changes and material flow impact. The formula for the pressure change over time in the S-shaped gradual rise curve is as follows: in the formula express The pressure value at any moment, Indicates the target pressure setpoint. Indicates the pressure build-up time. Indicates the duration of acceleration and deceleration. This indicates the total pressure build-up time. During the pressure build-up process, if a single station outputs an adjustment control quantity, the hydraulic cushion pressure of the adjacent tensioning stations is compensated according to the pre-stored adjacent crosstalk compensation coefficient. The adjacent crosstalk compensation coefficient is calibrated based on the elastic deformation test of the mold base. When the pressure of a single station is adjusted, the elastic deformation of the mold base will cause a slight change in the pressure of adjacent stations. A compensation quantity opposite to the direction of the adjustment quantity is added to the adjacent stations in advance to offset the pressure crosstalk caused by the elastic deformation of the mold base. The crosstalk compensation quantity is calculated using the following formula: in the formula This indicates the crosstalk compensation pressure that needs to be superimposed between adjacent workstations. This represents the adjacent crosstalk compensation coefficient, a dimensionless constant calibrated based on the mode base stiffness. This indicates the pressure adjustment amount for a single workstation. Crosstalk compensation can control the pressure fluctuation of adjacent workstations within 0.02 MPa when adjusting a single workstation, thus preventing the adjustment of a single workstation from affecting the normal molding of other workstations.

[0098] After entering the pressure holding and shaping stage, the pressure of the hydraulic cushion actuators at each station is kept stable. The control signals of the main cylinder electro-hydraulic proportional pressure valve and the hydraulic cushion actuators at each station are output synchronously. The time error of the control signal output does not exceed one control cycle, ensuring that the main punching pressure and the blank holder force at each stretching station reach the set value synchronously, and avoiding material flow loss of control caused by the asynchronous main pressure and blank holder force.

[0099] During the pressure holding process, if pressure fluctuations occur at a single station, the hydraulic cushion actuator at that station should be adjusted first. For overall pressure fluctuations, the electro-hydraulic proportional pressure valve of the main cylinder should be adjusted. Hysteresis compensation and crosstalk compensation should also be incorporated into the adjustment process. When the pressure holding ends and the pressure relief return phase begins, the hydraulic cushion actuators at each station are driven to depressurize in stages according to the pre-stored depressurization step size. In the initial stage of depressurization, a small step size is used for slow depressurization to release the elastic deformation energy of the material. In the middle stage of depressurization, a large step size is used for rapid depressurization to improve efficiency. In the final stage of depressurization, a small step size is used until the pressure reaches zero to avoid sudden depressurization causing damage to the workpiece. The elastic rebound deformation of the workpiece is graded, and the pressure relief step length is calibrated according to the elastic modulus of the material. For materials with high elasticity such as stainless steel, the pressure relief step length is smaller and the number of stages is more. For materials with low elasticity such as low carbon steel, the pressure relief step length is larger and the number of stages is less. For the hydraulic cushion actuator of the rigid impact process station, no pressure regulation control signal is input. The hydraulic cushion is kept in a fixed lower position throughout the stamping process. Only after the slide returns to the lower stop position of the slide does it perform a fixed stroke ejection action. The ejection action rises and ejects the workpiece according to a fixed speed curve. After the ejection is in place, it waits for the next stamping to start descending and resetting.

[0100] All drive signals are processed by a second-order low-pass filter before output. The filter cutoff frequency is set to 100 Hz to eliminate valve orifice jitter caused by high-frequency interference. After filtering, a small chatter signal with an amplitude of 0.5% of the rated current and a frequency of 100 Hz is superimposed on the drive signal to overcome the static friction of the proportional valve core and reduce the influence of valve hysteresis. The chatter signal amplitude is very small and will not cause observable pressure fluctuations.

[0101] After the drive signal is output, the actual displacement and pressure of the hydraulic cushion actuators at each station are continuously monitored. If the displacement exceeds the tolerance or the pressure is abnormal, the drive signal is immediately cut off, and an emergency stop signal is sent to protect the equipment and mold. During cold start-up, the hydraulic oil viscosity is high, so the hysteresis compensation is automatically increased, and the total pressure build-up time of the S-shaped gradual rise curve is appropriately extended to ensure a smooth pressure rise. As the equipment runs continuously, the oil temperature rises and the viscosity decreases, so the hysteresis compensation is automatically reduced, and the pressure build-up time is appropriately shortened to improve production efficiency. When adjusting a single station, crosstalk compensation is automatically triggered to avoid affecting adjacent stations. The step size and number of stages for graded pressure relief are automatically adjusted according to the material. This drive method, which combines viscosity adaptive hysteresis compensation, S-shaped gradual pressure build-up, crosstalk compensation, and graded pressure relief, can ensure stable material flow and improve the quality of workpiece forming.

[0102] Please see Figure 2 As shown, the second objective of this invention is to provide a method for implementing an adaptive control system for the punching force of a multi-station stretching punch press, comprising the following steps: S1. Real-time acquisition of slider-crank angle signal, identification of non-stamping idle window during slider return phase based on slider-crank angle signal, and outputting identification trigger signal; S2. After receiving the identification trigger signal, control the main cylinder unloading valve to open to cut off the pressure path of the main actuator, control the bypass solenoid valve to open the bypass branch where the fixed throttle is located, input a step-type identification current sequence to the electro-hydraulic proportional pressure valve, collect multiple sets of pressure difference values ​​at both ends of the fixed throttle under different steady-state flow rates, remove the inherent pressure pulsation component of the hydraulic pump and then invert to obtain the actual kinematic viscosity value of the current hydraulic oil, and immediately close the bypass solenoid valve after the identification is completed. S3. Compare the actual kinematic viscosity value with the reference viscosity value obtained and stored by the same identification process at the standard working oil temperature during the equipment's factory commissioning stage. Query the pre-stored relationship between viscosity and valve orifice flow gain to obtain the internal leakage compensation amount and control current bias amount of the electro-hydraulic proportional pressure valve at the current viscosity. S4. The blank holder force setting value and the main punch pressure setting value of each stretching station are superimposed with the corresponding control current bias value according to the pressure range of each station. At the same time, the current feedforward amount in the pressure building stage and the current maintenance amount in the pressure holding stage are corrected according to the internal leakage compensation amount. Combined with the pre-pressure building trigger logic, the corrected pressure setting value for the whole stroke is generated. S5. During the stamping stroke, the actual pressure value of the die holder at each station is collected by the pressure sensor at each station. The current stamping stroke stage is determined based on the slider crank angle signal. The actual pressure value of the die holder is compared with the corrected pressure set value of the corresponding station. The adjustment control quantity is calculated by combining the stage gain and control dead zone rules. The adjustment control quantity is output to distinguish between single station off-center load adjustment and overall pressure adjustment. S6. The corrected pressure setpoint and adjustment control quantity are combined with the valve characteristics under the current viscosity and converted into a current domain control signal. The hydraulic cushion actuators of each station are driven in the following manner: S-shaped slow pressure building in the pressure building stage, synchronous pressure of the main cylinder and hydraulic cushion in the pressure holding stage, and graded pressure relief in the pressure relief stage. For the rigid impact process station, only a fixed stroke ejection action is performed without pressure adjustment. After the pressure adaptive adjustment of the current stamping cycle is completed, it returns to S1.

[0103] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the foregoing description; the foregoing description and specifications are merely preferred embodiments and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. An adaptive control system for punching pressure of a multi-station stretching punch press, characterized in that, include: The stamping condition timing identification unit is used to collect the slider-crank angle signal in real time, identify the non-stamping idle window during the slider return phase based on the slider-crank angle signal, and output the identification trigger signal. The online identification unit for actual hydraulic viscosity receives the identification trigger signal and controls the opening of the master cylinder unloading valve to cut off the pressure path of the main actuator. It controls the bypass solenoid valve to open the bypass branch where the fixed throttle is located, inputs a stepped identification current sequence to the electro-hydraulic proportional pressure valve, collects multiple sets of pressure difference values ​​at different steady-state flow rates across the fixed throttle, removes the inherent pressure pulsation component of the hydraulic pump, and then inverts to obtain the actual kinematic viscosity value of the current hydraulic oil. After the identification is completed, the bypass solenoid valve is immediately closed. The valve control characteristic offset mapping unit is used to compare the actual kinematic viscosity value with the reference viscosity value obtained and stored by the same identification process at the standard working oil temperature during the equipment's factory commissioning stage. It queries the pre-stored correspondence between viscosity and valve orifice flow gain to obtain the internal leakage compensation amount and control current bias amount of the electro-hydraulic proportional pressure valve at the current viscosity. The pressure setpoint feedforward correction unit is used to superimpose the blank holder force setpoint of each stretching station with the main punch pressure setpoint and the corresponding control current bias. At the same time, it corrects the current feedforward during the pressure build-up stage and the current maintenance during the pressure holding stage based on the internal leakage compensation, and generates the corrected pressure setpoint. The stamping pressure closed-loop fine-tuning unit is used to collect the actual pressure value of the die holder during the stamping stroke, compare it with the corrected pressure set value, and output the adjustment control quantity to drive the hydraulic cushion actuator of each station to complete the adaptive adjustment of the stamping pressure.

2. The adaptive control system for punching pressure of a multi-station stretching punch press according to claim 1, characterized in that: The specific process for identifying the non-stamping idle window during the slider return phase based on the slider crank angle signal is as follows: The slider-crank rotation angle signal is acquired in real time. The bottom dead center position and top dead center position of the slider are determined based on the slider-crank rotation angle signal. The stroke range of the slider from the bottom dead center position to the top dead center position is determined as the slider return range. Within the slide return range, the residual pressure of the hydraulic system is detected by the main cylinder pressure sensor, and the workpiece transfer completion signal output by the multi-station feeding robot controller is received at the same time. When the position corresponding to the slider crank angle signal crosses the preset angle before the top dead center of the slider, the residual pressure of the hydraulic system is lower than the preset pressure threshold, and the workpiece transmission completion signal is received, it is determined that the non-stamping idle window of the slider return stage is entered and an identification trigger signal is output. The slider crank angle signal is monitored in real time. When the remaining time between the slider and the next stamping contact with the workpiece is less than the minimum time required for viscosity identification, the non-stamping idle window of the slider return phase is ended in advance. The minimum time required for viscosity identification is the fixed time required for the online identification unit of the actual viscosity of the oil to complete one complete identification process.

3. The adaptive control system for punching pressure of a multi-station stretching punch press according to claim 1, characterized in that: The stepped identification current sequence specifically includes: The dead zone current of the electro-hydraulic proportional pressure valve is retrieved from the valve control characteristic offset mapping unit. The dead zone current of the electro-hydraulic proportional pressure valve is used as the starting current value of the step-type identification current sequence. The dead zone current of the electro-hydraulic proportional pressure valve is the minimum operating current calibrated and stored at the factory of the electro-hydraulic proportional pressure valve. The output current is increased step by step according to a pre-stored fixed current step size. Each current plateau is maintained for a pre-stored steady-state holding time. The steady-state holding time is a fixed time preset according to the hydraulic system pressure response time constant to wait for the pressure difference between the two ends of the fixed throttle to reach a steady state. A pressure stability determination window is set at the end of each current platform. The pressure difference value across the fixed throttle is collected only within the pressure stability determination window to avoid the hydraulic shock fluctuations caused by the instantaneous current step. The pressure stability determination window is a fixed time window at the end of each current platform for collecting steady-state pressure difference values. The highest current value of the stepped identification current sequence corresponds to the median range of the rated operating current of the electro-hydraulic proportional pressure valve, and the total execution time of the entire stepped identification current sequence does not exceed the shortest time required for viscosity identification. After collecting differential pressure values ​​for all stages, first reduce the output current of the electro-hydraulic proportional pressure valve to zero, and then close the bypass solenoid valve.

4. The adaptive control system for punching pressure of a multi-station stretching punch press according to claim 1, characterized in that: The specific process of obtaining the actual kinematic viscosity value of the hydraulic oil after eliminating the inherent pressure pulsation component of the hydraulic pump is as follows: Extract the pressure difference values ​​across all fixed throttles collected within the pressure stability determination window, and extract the pressure difference value sequence by whole cycle according to the pre-stored inherent pulsation cycle of the hydraulic pump. The inherent pulsation cycle of the hydraulic pump is a fixed cycle value calibrated and stored at the factory based on its rated speed and number of plungers. The pressure difference value within each complete hydraulic pump inherent pulsation cycle is arithmetically averaged. The periodic hydraulic pump inherent pressure pulsation component is eliminated due to positive and negative cancellation, resulting in the steady-state pressure difference value of each current level in the corresponding step-type identification current sequence. Remove abnormal data points that deviate from the average steady-state differential pressure value of the same level by more than a preset threshold, eliminate random errors, and retrieve the calibration differential pressure values ​​corresponding to each level of current obtained by the same step-by-step identification current sequence calibration at standard working oil temperature during the equipment's factory commissioning stage stored in the valve control characteristic offset mapping unit. The steady-state voltage difference value corresponding to each current level is compared with the calibrated voltage difference value under the same current to obtain the voltage difference deviation of each current level. Based on the pre-stored correspondence between fixed throttle flow rate, pressure difference, and viscosity, the viscosity deviation value of each level is calculated from the pressure difference deviation of each level. Then, the viscosity deviation values ​​of each level are weighted according to the pre-stored weighted average coefficient. Combined with the reference viscosity value, the actual kinematic viscosity value of the current hydraulic oil is obtained. The weighted average coefficient is the pre-stored weight value corresponding to different current ranges.

5. The adaptive control system for punching pressure of a multi-station stretching punch press according to claim 1, characterized in that: The specific relationship between viscosity and valve orifice flow gain includes: Multiple standard working oil temperature points are set up to cover the entire temperature range from cold start-up to continuous operation. At each standard working oil temperature point, the same step-type identification current sequence and pressure difference value processing procedure as the online identification is adopted to test and obtain the complete valve control characteristic curve family of the electro-hydraulic proportional pressure valve at each calibrated viscosity. The calibrated viscosity is the kinematic viscosity value of hydraulic oil measured by the same identification procedure at each standard working oil temperature point. The viscosity measured at the standard working oil temperature reference point is the reference viscosity value. The viscosity-valve flow gain correspondence relationship includes the pre-stored zero-point leakage current value at each calibrated viscosity, the valve flow gain coefficient across the entire pressure range, the pressure build-up feedforward coefficient, and the pressure maintenance coefficient. The valve flow gain coefficient is the output pressure change rate of the electro-hydraulic proportional pressure valve corresponding to a unit change in control current. When querying online, the actual kinematic viscosity of the hydraulic oil is used as the index, and linear interpolation is used between two adjacent calibrated viscosity points to obtain the zero leakage current value at the current viscosity and the valve orifice flow gain coefficient for each pressure range. The zero-point leakage current value is combined with the pressure build-up feedforward coefficient and the pressure holding maintenance coefficient to convert it into an internal leakage compensation amount, which is used to correct the current feedforward amount in the pressure build-up stage and the current holding amount in the pressure holding stage. The difference between the valve orifice flow gain coefficient at each pressure range under the current viscosity and the reference flow gain coefficient under the reference viscosity is converted into a control current bias amount, which is used to superimpose the blank holder force setting value and the main punch pressure setting value at each stretching station.

6. The adaptive control system for punching pressure of a multi-station stretching punch press according to claim 1, characterized in that: The specific process of superimposing the blank holder force setting value and the main punch pressure setting value of each stretching station with the corresponding control current bias is as follows: The complete stamping stroke is divided into material contact, pressure building and stretching, pressure holding and shaping, and pressure release and return stroke according to the slider crank rotation angle signal. The blank holder force setting value and the main punch pressure setting value of each stretching station are mapped to the pre-stored pressure range mapping table according to the pressure magnitude. The pressure range mapping table is a pressure segment correspondence table that corresponds one-to-one with the flow gain coefficient of each pressure range valve port. For each stretching station, the blank holder force setting value is independently superimposed according to the pressure range in which its own pressure magnitude is located. Different stretching stations are matched with the corresponding control current bias value according to the range in which their own blank holder force setting value is located. For the main punch pressure setting value that changes dynamically with the slider crank angle signal, the control current bias of the main punch pressure in the pressure range at the current stroke position is matched point by point according to the fixed crank angle step and then superimposed. The pressure setting value of the rigid impact process station performing edge trimming, punching, and blanking is not superimposed with the control current bias to avoid overcompensation under rigid impact conditions. After the bias of the blank holder force setting value and the main punch pressure setting value of all stretching stations are superimposed, the final corrected pressure setting value is generated by combining the current feedforward amount of the pressure building stage and the current maintenance amount of the pressure holding stage after the correction based on the internal leakage compensation amount.

7. The adaptive control system for punching pressure of a multi-station stretching punch press according to claim 1, characterized in that: The specific process of simultaneously correcting the current feedforward during the pressure build-up phase and the current maintenance during the pressure holding phase based on the internal leakage compensation to generate the corrected pressure setpoint is as follows: The pre-pressure trigger position is determined based on the slider crank angle signal. The pre-pressure trigger position is the position where the crank angle is fixed before the slider reaches the material contact stage. This position is a fixed position calibrated at the factory according to different stamping strokes. Before entering the pressure building and stretching stage, the current feedforward amount for the pressure building stage is injected from the pre-pressure building trigger position. The current feedforward amount for the pressure building stage is obtained by multiplying the internal leakage compensation amount by the pressure building feedforward coefficient of the corresponding station, which is used to compensate for the pressure building lag caused by the internal leakage of the electro-hydraulic proportional pressure valve and the cylinder. Different stretching stations pre-store corresponding pressure build-up feedforward coefficients according to process requirements. The pressure build-up feedforward coefficient of the first deep stretching station is higher than that of the subsequent shaping stations. After entering the pressure holding and shaping stage, the current holding amount of the pressure holding stage is continuously injected. The current holding amount of the pressure holding stage is obtained by multiplying the internal leakage compensation amount by the pressure holding coefficient of the corresponding station, and is used to compensate for the pressure drop caused by internal leakage during the pressure holding process. Different stretching stations pre-store corresponding pressure holding coefficients according to the pressure holding time requirements. For rigid impact process stations, the current feedforward amount during the pressure building stage and the current holding amount during the pressure holding stage are not injected. After superimposing the control current bias of the blank holder force setting value and the main punch pressure setting value at each stretching station, and correcting the current feedforward amount in the pressure building stage and the current maintenance amount in the pressure holding stage, the corrected pressure setting value for the entire stroke is generated.

8. The adaptive control system for punching pressure of a multi-station stretching punch press according to claim 1, characterized in that: The specific process of acquiring the actual pressure value of the die holder during the stamping stroke, comparing it with the corrected pressure setpoint, and outputting the adjustment control is as follows: The actual pressure value of the die holder at the corresponding workstation is collected by the pressure sensor at the workstation. The current stamping stroke stage is determined based on the slider crank rotation angle signal, and the pre-stored stage gain coefficient of the corresponding stage is called. The actual pressure value of the die holder at each stretching station is compared with the corrected pressure setting value of the corresponding station point by point to calculate the pressure deviation value. When the pressure deviation value is within the pre-stored control dead zone threshold range, no adjustment control quantity is output to avoid frequent operation of the electro-hydraulic proportional pressure valve and pressure oscillation. When the pressure deviation value of a single stretching station exceeds the pre-stored off-center load deviation threshold, an adjustment control quantity is output separately to the hydraulic cushion actuator of the corresponding station to compensate for the single-station pressure deviation caused by off-center load of multiple stations. When the overall pressure deviation of all stations is consistent in both direction and magnitude, the output adjustment control quantity is sent to the electro-hydraulic proportional pressure valve of the main cylinder. For the rigid impact process station, only the actual pressure value of the mold base is collected for over-limit shutdown protection. No adjustment control quantity is output. The output adjustment control quantity is superimposed on the feedforward corrected control quantity to drive the hydraulic cushion actuator of each station to complete the adaptive adjustment of the impact pressure.

9. The adaptive control system for punching pressure of a multi-station stretching punch press according to claim 1, characterized in that: The specific process by which the hydraulic cushion actuators at each workstation complete the adaptive adjustment of the punching force is as follows: The corrected pressure setpoint and adjustment control quantity in the pressure domain are combined with the characteristics of the electro-hydraulic proportional pressure valve under the current viscosity and converted into a current domain control signal. The conversion process uses the valve orifice flow gain and internal leakage compensation parameters under the current viscosity output by the valve control characteristic offset mapping unit, and does not use the fixed conversion coefficient under the reference viscosity. During the pressure building and stretching stage, the hydraulic cushion actuators at each station are driven to build pressure according to the pre-stored S-shaped gradual rise curve. The pressure change rate does not exceed the pre-stored pressure change rate limit value to avoid sudden pressure changes that could cause material flow impact. When a single station outputs an adjustment control quantity, the hydraulic cushion pressure of the adjacent two stretching stations is compensated according to the pre-stored adjacent crosstalk compensation coefficient to offset the pressure crosstalk caused by the elastic deformation of the mold base. During the pressure holding and shaping stage, the pressure of the hydraulic cushion actuators at each station is kept stable. The control signals of the main cylinder electro-hydraulic proportional pressure valve and the hydraulic cushion actuators at each station are output synchronously to ensure that the main punching pressure and the blank holder force at each stretching station reach the set value synchronously. When the pressure holding period ends and the pressure relief return phase begins, the hydraulic cushion actuators at each station are driven to release pressure in stages according to the pre-stored graded pressure relief step length, so as to avoid sudden pressure relief causing elastic rebound deformation of the workpiece. For the hydraulic cushion actuator of the rigid impact process station, no pressure regulation control signal is input. It only performs a fixed stroke ejection action after the slider returns to the lower stop position of the slider. All drive signals are low-pass filtered before output to eliminate valve port jitter caused by high frequency interference.

10. A method for implementing an adaptive control system for the punching force of a multi-station stretching punch press as described in any one of claims 1-9, characterized in that: Includes the following steps: S1. Real-time acquisition of slider-crank angle signal, identification of non-stamping idle window during slider return phase based on slider-crank angle signal, and outputting identification trigger signal; S2. After receiving the identification trigger signal, control the main cylinder unloading valve to open to cut off the pressure path of the main actuator, control the bypass solenoid valve to open the bypass branch where the fixed throttle is located, input a step-type identification current sequence to the electro-hydraulic proportional pressure valve, collect multiple sets of pressure difference values ​​at both ends of the fixed throttle under different steady-state flow rates, remove the inherent pressure pulsation component of the hydraulic pump and then invert to obtain the actual kinematic viscosity value of the current hydraulic oil, and immediately close the bypass solenoid valve after the identification is completed. S3. Compare the actual kinematic viscosity value with the reference viscosity value obtained and stored by the same identification process at the standard working oil temperature during the equipment's factory commissioning stage. Query the pre-stored relationship between viscosity and valve orifice flow gain to obtain the internal leakage compensation amount and control current bias amount of the electro-hydraulic proportional pressure valve at the current viscosity. S4. The blank holder force setting value and the main punch pressure setting value of each stretching station are superimposed with the corresponding control current bias value according to the pressure range of each station. At the same time, the current feedforward amount in the pressure building stage and the current maintenance amount in the pressure holding stage are corrected according to the internal leakage compensation amount. Combined with the pre-pressure building trigger logic, the corrected pressure setting value for the whole stroke is generated. S5. During the stamping stroke, the actual pressure value of the die holder at each station is collected by the pressure sensor at each station. The current stamping stroke stage is determined based on the slider crank angle signal. The actual pressure value of the die holder is compared with the corrected pressure set value of the corresponding station. The adjustment control quantity is calculated by combining the stage gain and control dead zone rules. The adjustment control quantity is output to distinguish between single station off-center load adjustment and overall pressure adjustment. S6. The corrected pressure setpoint and adjustment control quantity are combined with the valve characteristics under the current viscosity and converted into a current domain control signal. The hydraulic cushion actuators of each station are driven in the following manner: S-shaped slow pressure building in the pressure building stage, synchronous pressure of the main cylinder and hydraulic cushion in the pressure holding stage, and graded pressure relief in the pressure relief stage. For the rigid impact process station, only a fixed stroke ejection action is performed without pressure adjustment. After the pressure adaptive adjustment of the current stamping cycle is completed, it returns to S1.

Images