Mould locking return-to-zero control system and method for injection molding machine

Abstract

The invention relates to the technical field of injection molding machine mold locking return-to-zero control, and discloses an injection molding machine mold locking return-to-zero control system and method. Meanwhile, temperature sensors are arranged on the pull rod, the movable template and the guide rail, and thermal expansion amount of each component is calculated and dynamically compensated in real time; calculating the wear compensation amount based on the accumulated number of cycles of the guide rail; fusing the position of the encoder, temperature compensation and wear compensation to obtain a real position; when approaching a mechanical zero point, monitoring a zero point sensor signal which is subjected to wavelet filtering and self-adaptive threshold processing; and finally, the driving mechanism moves to the zero point in a closed loop mode and automatically stops within the error controllable range.

Description

Technical Field

[0001] The present invention relates to the technical field of injection molding machine clamping zeroing control, and specifically to an injection molding machine clamping zeroing control system and method. Background Art

[0002] As the core equipment for plastic product production in modern manufacturing, the accuracy and stability of the clamping mechanism of an injection molding machine directly affect product quality and production efficiency. Clamping zeroing control is a key step after the injection molding machine is started or the mold is changed, aiming to accurately reset the moving platen to the mechanical origin to ensure the repeatability accuracy of subsequent clamping and injection operations. However, the traditional clamping zeroing control technology has the following significant problems:

[0003] When the injection molding machine operates for a long time or the ambient temperature changes, the metal components of the clamping mechanism deform due to thermal expansion and contraction; the traditional method does not perform dynamic compensation for this, resulting in the actual position deviating from the designed zero point after zeroing, affecting the molding quality of injection molded parts; components such as guide rails and lead screws are locally worn after long-term high-frequency operation, resulting in errors in the clamping mechanism due to wear, which will also lead to the accuracy of clamping zeroing and affect the molding quality of injection molded parts; in addition, electromagnetic noise generated by equipment such as inverters and high-power motors in the injection molding machine workshop is coupled into the zero sensor signal, resulting in interference of the sensor signal and thus the risk of false triggering;

[0004] Based on this, this solution proposes a multi-modal adaptive compensation technology to achieve high-precision and low-cost clamping zeroing control. Summary of the Invention

[0005] The present invention provides an injection molding machine clamping zeroing control system and method, which helps to solve the problems mentioned in the above background art.

[0006] The present invention provides the following technical solution: an injection molding machine clamping zeroing control method, including:

[0007] Taking 10 ms as a cycle, and performing clamping zeroing calibration control within each cycle, specifically:

[0008] Obtaining the real-time position of the clamping mechanism through an encoder, denoted as P encoder ;

[0009] Collecting the temperatures of the tie rods, platens, and guide rails through a temperature detection module, and respectively denoting them as T1, T2, and T3;

[0010] Calculating the thermal expansion amounts of each component according to the temperatures of the tie rods, platens, and guide rails, and taking the obtained thermal expansion amounts as dynamic temperature compensation amounts;

[0011] Calculating the movement trajectory deviation caused by mechanical wear according to the number of cycles of the current injection molding machine guide rail working, as the wear compensation amount;

[0012] Based on the real-time position P of the die-locking mechanism encoder and combining the dynamic temperature compensation amount and the wear compensation amount, the real position of the die-locking mechanism is obtained;

[0013] According to the real position of the die-locking mechanism, control the die-locking mechanism to move towards the mechanical zero point;

[0014] During the process that the die-locking mechanism approaches the mechanical zero point, monitor the signal state of the zero point sensor;

[0015] Filter the signals of the monitored zero point sensor and retain the real zero point sensor signals;

[0016] Drive the die-locking mechanism to move until it returns to the mechanical zero point, completing the zeroing of the die-locking mechanism.

[0017] Optionally, calculating the thermal expansion amounts of each component according to the temperatures of the tie rod, the template, and the guide rail, and using the obtained thermal expansion amounts as the dynamic temperature compensation amount includes:

[0018] Calculating the thermal expansion amounts of the tie rod, the template, and the guide rail:

[0019]

[0020] Wherein:

[0021] α i : The linear expansion coefficient of the tie rod, the template, and the guide rail, obtained according to the material of each component;

[0022] L i : The initial lengths of the tie rod, the template, and the guide rail, with the unit of mm;

[0023] T ref : The reference temperature, set to 20 °C;

[0024] β: The dynamic compensation coefficient, calibrated through a thermal shock experiment;

[0025] T i : The temperatures of the tie rod, the template, and the guide rail are collected by temperature sensors, which are T1, T2, and T3 respectively;

[0026] The temperature change rate is obtained by calculating the temperature difference between adjacent cycles:

[0027]

[0028] Wherein:

[0029] T i (t): The temperature of this cycle;

[0030] T i(t - 1): Temperature of the previous cycle;

[0031] Then the position data after removing the thermal expansion amount is denoted as P DTC ;

[0032] P DTC = P encoder -ΔL·γ

[0033] Where:

[0034] γ: Represents the temperature - position factor, obtained by calculating from historical data.

[0035] Optionally, the calibration is performed through a thermal shock experiment, specifically:

[0036] Perform M experiments on the mode - locking mechanism, and apply temperature steps at different rates in each experiment;

[0037] In each experiment, record the temperature change rate and the actual expansion amount ΔL real,k , and fit β by the least - squares method, take the derivative of the error function and set the derivative to zero:

[0038]

[0039] Where:

[0040] M: Represents the total number of samples of the experimental data, that is, the number of thermal shock experimental data points used to calibrate β;

[0041] ΔL real,k : The measured thermal expansion amount in the k - th experiment, with the unit of mm;

[0042] ΔT k : The difference between the temperature collected and the reference temperature in the k - th experiment;

[0043] The temperature change rate of the k - th experiment.

[0044] Optionally, the obtaining through historical data calculation includes:

[0045] Under the no - load state of the device, collect V groups of temperature, encoder position, and real position data calibrated manually;

[0046] Calculate ΔL for the k - th group of data k ;

[0047] Calculate the temperature - position factor as:

[0048]

[0049] Where:

[0050] P encoder,k : Encoder position measured in the k-th group of data, unit: mm;

[0051] P real,k : True position manually calibrated in the k-th group of data, unit: mm;

[0052] ΔL k : Calculated thermal expansion in the k-th group of data, unit: mm;

[0053] V: Number of groups of collected data.

[0054] Optionally, calculating the motion trajectory deviation caused by mechanical wear according to the number of cycles of the current injection molding machine guide rail as the wear compensation amount, including:

[0055] Denote the wear compensation amount as ΔP WCM , then

[0056] ΔP WCM = W guide ·(a·P DTC + b)

[0057] Where:

[0058] W guide : Wear coefficient, obtained by calculating the number of cycles;

[0059] a, b: Linear coefficients, obtained by regression calculation of historical data;

[0060] The obtaining by calculating the number of cycles is specifically:

[0061] W guide = s·X

[0062] X: Number of cycles of the current injection molding machine guide rail;

[0063] s: Wear coefficient, obtained according to the statistical result of the industry average wear rate of the injection molding machine guide rail.

[0064] Optionally, the linear coefficients are obtained by regression calculation of historical data, specifically:

[0065] Obtain Z groups of historical data, each group of data includes the position P after temperature compensation DTC,k , actual wear deviation ΔP real,k , wear coefficient W guide,k ;

[0066] Fitting through historical data:

[0067] a represents the deviation amount caused by the position increasing by 1 mm per unit wear coefficient, and the calculation formula is:

[0068]

[0069] b reflects the base wear offset independent of position, and the calculation formula is:

[0070]

[0071] Where:

[0072] P DTC,k : represents the calculated P in the k-th group of data DTC ;

[0073] ΔP real,k : represents the difference between the true position measured by the dial indicator and the encoder position in the k-th group of data;

[0074] W guide,k : represents the calculated W in the k-th group of data guide .

[0075] Optionally, based on the real-time position P of the mold clamping mechanism encoder , combining the dynamic temperature compensation amount and the wear compensation amount, to obtain the real position of the mold clamping mechanism, including:

[0076] P target = P DTC + ΔP WCM

[0077] Where: P target is the corrected real position, with the unit of mm.

[0078] Optionally, filtering the signals of the monitored zero-point sensor to retain the real zero-point sensor signals, including:

[0079] Collect the signals of the zero-point sensor and denote it as S raw ;

[0080] Perform anti-interference filtering on S raw to obtain the filtered signal S filtered :

[0081] S filtered = IDWT(SoftThresholding(DWT(S raw ), λ))

[0082] Where:

[0083] DWT: Discrete Wavelet Transform, using the db4 wavelet basis;

[0084] SoftThresholding: Soft threshold processing function;

[0085] IDWT: Inverse Discrete Wavelet Transform, reconstructing the filtered signal;

[0086] λ: Dynamic threshold, adaptively adjusted according to the interference frequency, specifically:

[0087]

[0088] Wherein:

[0089] λ0: Basic threshold, the reference of the sensor noise level; the reference of the background noise amplitude of the sensor when there is no effective signal, used to distinguish the real signal from the noise; when the mode-locking mechanism is stationary (no movement, no load), continuously collect the output signal of the sensor and calculate the root mean square value of the signal as the basic threshold;

[0090] f: Current signal frequency component;

[0091] f noise : Main interference frequency; by collecting the sensor signal during the normal operation of the device, performing a fast Fourier transform on the signal, identifying the high-frequency peaks with significant amplitudes, and recording the frequency corresponding to the peak as the main interference frequency;

[0092] σ: Frequency bandwidth coefficient; specifically, by collecting multiple groups of signals containing interference, analyzing the distribution range of the interference frequency through a fast Fourier transform, and calculating the standard deviation of the frequency distribution as the frequency bandwidth coefficient;

[0093] If within three consecutive cycles, S filtered > λ, it is determined that the real mechanical zero point is detected.

[0094] Optionally, driving the mode-locking mechanism to move until it returns to the mechanical zero point to complete the zeroing of the mode-locking mechanism includes:

[0095] Setting the maximum allowable zeroing error value, denoted as ∈;

[0096] In each cycle, perform position error calculation:

[0097] μ = P target - P encoder

[0098] Wherein:

[0099] P target : The true position of the mode-locking mechanism obtained by combining the dynamic temperature compensation amount and the wear compensation amount within the cycle;

[0100] P encoder : The position of the mode-locking mechanism displayed by the encoder within the cycle;

[0101] When a true mechanical zero point is detected and |μ| < ∈, the control of the mold clamping mechanism is stopped, and the mold clamping mechanism is reset to zero.

[0102] A system for implementing the mold clamping zeroing control method of the injection molding machine includes:

[0103] Encoder module: Displays the position of the mold clamping mechanism on the mold clamping mechanism;

[0104] Temperature detection module: Used to collect the temperatures of the tie bars, templates, and guide rails;

[0105] Zero point sensor module: Used to detect and identify the mechanical zero point;

[0106] Main controller module: Executes the multi-modal compensation algorithm;

[0107] Signal filtering module: Filters the signals collected by the zero point sensor module;

[0108] Clock synchronization module: Provides a unified time reference for the system to ensure the synchronization of data acquisition and control cycles;

[0109] Drive module: Drives the movement of the mold clamping mechanism.

[0110] Data storage module: Records historical data for regression calculation of parameters.

[0111] The present invention has the following beneficial effects:

[0112] 1. By calculating the thermal expansion amount based on the temperatures of the tie bars, moving templates, and guide rails through the linear expansion coefficient and dynamic compensation coefficient, and performing real-time compensation, the position reduction accuracy can be significantly improved; in this step, first, according to the material linear expansion coefficient, reference temperature, and real-time temperature difference, combined with the dynamic compensation coefficient obtained by experimental calibration, the thermal expansion amounts of each component are accurately calculated; then, this expansion amount is used as the dynamic temperature compensation amount to remove the false displacement introduced by thermal deformation; when the temperature change rate is large, the compensation amount can be weighted and adjusted through the temperature difference between adjacent cycles, reflecting the "rate perception" characteristic; this method can better adapt to the working condition fluctuations compared with the traditional fixed-value compensation; the temperature-position factor obtained by historical data regression can further suppress the non-linear error; the real-time compensation scheme avoids the mold clamping misalignment and flash defects caused by temperature drift, and ensures the geometric tolerance and surface quality of the injection molded parts.

[0113] 2. Calculating the mechanical wear coefficient by accumulating the number of guide rail cycles and generating the wear compensation amount can extend the effective life of the injection molding machine and ensure long-term positioning accuracy. In this step, the number of guide rail cycles is combined with the industry average wear rate statistics to obtain the real-time wear coefficient. Subsequently, using the linear coefficient and the basic offset obtained by historical data regression calculation, the wear coefficient is converted into a displacement compensation amount. This compensation amount is used to correct the trajectory deviation caused by the wear of the guide rail screw or roller due to long-term operation. Compared with regular manual calibration or component replacement, online wear compensation can delay the hardware replacement cycle and reduce maintenance costs. In the scenario of multi-shift and high-load continuous operation of the equipment, wear compensation can smoothly eliminate the accumulated error and maintain the mold closing and zeroing accuracy at the micron level. At the same time, this solution can track the wear trend in real time, support maintenance prediction and spare parts management. Combining with thermal expansion compensation to form a multi-modal adaptive correction ensures the long-term stability of the zeroing accuracy.

[0114] 3. Fusing the real-time encoder position, dynamic temperature compensation amount and wear compensation amount to calculate the real position effectively solves the positioning deviation caused by the superposition of multi-source errors. In this step, the three are aggregated according to the weighted or superposition model to obtain the corrected real position. This fusion algorithm is implemented at the software level without additional sensors or machining parts. The fused real position not only eliminates the temperature drift error and wear offset, but also filters out occasional measurement noise. On this basis, performing a zeroing movement can ensure that the movement end point is highly consistent with the mechanical origin. This method can be implemented by firmware or PLC upgrade on existing injection molding machines, with the advantages of fast deployment and upgrade iteration. The accurate real position feedback can also provide reliable data for the upper-level MES or CPS system, promoting intelligent manufacturing.

[0115] 4. Performing wavelet anti-interference filtering on the zero-point sensor signal and using an adaptive dynamic threshold to retain the real signal significantly improves the signal reliability. In this step, the collected discrete signal is first wavelet decomposed, and the soft threshold processing is used to remove the high-frequency interference components. Then, the filtered signal is reconstructed through inverse transformation. The dynamic threshold is adaptively adjusted according to the main interference frequency and the frequency band width coefficient in the real-time signal spectrum, making the boundary between noise and signal more accurate. The real mechanical zero point can be determined only when the filtered signal continuously meets the conditions in three consecutive cycles. This multi-level filtering and threshold strategy has stronger robustness and anti-interference ability compared with the fixed threshold scheme. It can also stably detect the real zero-point signal in the strong electromagnetic noise and high-frequency vibration environment in the workshop, avoiding false touches and missed touches.

[0116] 5. The drive clamping mechanism moves towards the mechanical zero point according to the actual position and automatically stops in combination with the maximum allowable error, enabling precise zeroing. In this step, the error between the actual position and the encoder position is calculated in real time for each cycle, and the error is compared with the set maximum zeroing error value. When the actual mechanical zero point is detected and the error is within the allowable range, the drive is automatically stopped. If the error exceeds the limit, the system will move in the reverse direction or make a decelerated fine adjustment until the condition is met. This closed-loop iterative strategy avoids excessive impact and control oscillation, ensuring stable convergence of the clamping zeroing. BRIEF DESCRIPTION OF THE DRAWINGS

[0117] Figure 1 It is a schematic diagram of the basic process of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0118] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0119] Example 1. Referring to Figure 1 , an injection molding machine clamping zeroing control method, including:

[0120] Taking 10 ms as a cycle, clamping zeroing calibration control is performed within each cycle, specifically:

[0121] Obtain the real-time position of the clamping mechanism through the encoder, denoted as P encoder ;

[0122] Collect the temperatures of the tie bars, templates, and guide rails through the temperature detection module, and record them as T1, T2, and T3 respectively;

[0123] Calculate the thermal expansion amounts of each component according to the temperatures of the tie bars, templates, and guide rails, and use the obtained thermal expansion amounts as dynamic temperature compensation amounts;

[0124] Calculate the motion trajectory deviation caused by mechanical wear according to the number of cycles of the current injection molding machine guide rail working, as the wear compensation amount;

[0125] On the basis of the real-time position P encoder of the clamping mechanism, combine the dynamic temperature compensation amount and the wear compensation amount to obtain the real position of the clamping mechanism;

[0126] Control the clamping mechanism to move towards the mechanical zero point according to the real position of the clamping mechanism;

[0127] During the process of the clamping mechanism approaching the mechanical zero point, monitor the signal state of the zero point sensor;

[0128] Filter the signals of the monitored zero-point sensors and retain the true zero-point sensor signals;

[0129] Drive the mode-locking mechanism to move until it returns to the mechanical zero point, completing the zeroing of the mode-locking mechanism.

[0130] Calculating the thermal expansion amounts of each component based on the temperatures of the tie rod, template, and guide rail, and using the obtained thermal expansion amounts as dynamic temperature compensation amounts, includes:

[0131] Calculate the thermal expansion amounts of the tie rod, template, and guide rail:

[0132]

[0133] Wherein:

[0134] α i : The linear expansion coefficient of the tie rod, template, and guide rail, obtained corresponding to the material of each component;

[0135] L i : The initial lengths of the tie rod, template, and guide rail, with the unit of mm;

[0136] T ref : The reference temperature, set to 20 °C; the 20 °C is the standard measurement temperature of the material thermal expansion coefficient, set based on the ISO 1:2002 standard;

[0137] β: The dynamic compensation coefficient, calibrated through thermal shock experiments;

[0138] T i : Collect the temperatures of the tie rod, template, and guide rail through temperature sensors, which are T1, T2, and T3 respectively;

[0139] The temperature change rate is obtained by calculating the temperature difference between adjacent cycles:

[0140]

[0141] Wherein:

[0142] T i (t): The temperature of this cycle;

[0143] T i (t - 1): The temperature of the previous cycle;

[0144] Then the position data after removing the thermal expansion amount is denoted as P DTC ;

[0145] P DTC =P encoder -ΔL·γ

[0146] Wherein:

[0147] γ represents the temperature - position factor, which is obtained by calculating historical data. By calculating the thermal expansion amount based on the temperature of the tie rod, moving platen and guide rail through the linear expansion coefficient and dynamic compensation coefficient, and performing real - time compensation, the position reduction accuracy can be significantly improved. In this step, first, according to the linear expansion coefficient of the material, the reference temperature and the real - time temperature difference, combined with the dynamic compensation coefficient obtained by experimental calibration, the thermal expansion amount of each component is accurately calculated. Subsequently, this expansion amount is used as the dynamic temperature compensation amount to remove the false displacement introduced by thermal deformation. When the temperature change rate is large, the compensation amount can be weighted and adjusted through the temperature difference of adjacent cycles, reflecting the "rate perception" characteristic. This method can better adapt to the working condition fluctuations compared with the traditional fixed - value compensation. The temperature - position factor obtained by regression of historical data can further suppress the non - linear error. The real - time compensation scheme avoids the problems of mold - locking misalignment and flash defects caused by temperature drift, and ensures the geometric tolerance and surface quality of the injection - molded parts.

[0148] The calibration through the thermal shock experiment is specifically as follows:

[0149] Perform M experiments on the mold - locking mechanism, and apply temperature steps at different rates in each experiment;

[0150] In each experiment, record the temperature change rate and the actual expansion amount ΔL real,k , and fit β by the least - squares method, take the derivative of the error function and set the derivative to zero:

[0151]

[0152] Wherein:

[0153] M represents the total number of samples of the experimental data, that is, the number of thermal shock experimental data points used to calibrate β;

[0154] ΔL real,k : the measured thermal expansion amount in the k - th experiment, with the unit of mm;

[0155] ΔT k : the difference between the temperature collected in the k - th experiment and the reference temperature;

[0156] The temperature change rate of the k - th experiment.

[0157] The obtaining by calculating historical data includes:

[0158] Under the no - load state of the equipment, collect V groups of temperature, encoder position and real - position data calibrated manually;

[0159] Calculate ΔL for the k - th group of data k ;

[0160] The temperature-position factor is calculated as follows:

[0161]

[0162] Where:

[0163] P encoder,k : The encoder position measured in the k-th group of data, in mm;

[0164] P real,k : The true position manually calibrated in the k-th group of data, in mm;

[0165] ΔL k : The calculated thermal expansion in the k-th group of data, in mm;

[0166] V: The number of groups of collected data.

[0167] Calculating the motion trajectory deviation caused by mechanical wear based on the number of cycles of the current injection molding machine guide rail as the wear compensation amount, including:

[0168] Denote the wear compensation amount as ΔP WCM , then

[0169] ΔP WCM = W guide ·(a·P DTC + b)

[0170] Where:

[0171] W guide : The wear coefficient, obtained by calculating the number of cycles;

[0172] a, b: The linear coefficients, obtained by regression calculation of historical data;

[0173] The obtaining by calculating the number of cycles is specifically:

[0174] W guide = s·X

[0175] X: The number of cycles of the current injection molding machine guide rail;

[0176] s: The wear coefficient, obtained according to the statistical result of the industry average wear rate of the injection molding machine guide rail; for example, the life test of a certain brand of guide rail shows that the wear amount per million cycles is about 1 mm, then the wear rate Calculating the mechanical wear coefficient by accumulating the number of guide rail cycles and generating the wear compensation amount can extend the effective life of the injection molding machine and ensure long-term positioning accuracy; this step combines the number of guide rail cycle operations with the industry average wear rate statistics to obtain the real-time wear coefficient; subsequently, using the linear coefficient and the basic offset obtained by regression calculation of historical data, the wear coefficient is converted into a displacement compensation amount; this compensation amount is used to correct the trajectory deviation caused by the wear of the guide rail screw or roller due to long-term operation; compared with regular manual calibration or component replacement, on-line wear compensation can delay the hardware replacement cycle and reduce the maintenance cost; in the scenario of multi-shift and high-load continuous operation of the equipment, wear compensation can smoothly eliminate the accumulated error and maintain the mold clamping zero return at the micron level accuracy; at the same time, this solution can track the wear trend in real time, support maintenance prediction and spare parts management; combined with thermal expansion compensation to form multi-modal adaptive correction, ensuring the long-term stability of the zero return accuracy.

[0177] The linear coefficient is obtained through regression calculation of historical data, specifically:

[0178] Obtain Z groups of historical data, each group of data includes the position P after temperature compensation DTC,k , the actual wear deviation ΔP real,k , the wear coefficient W guide,k ;

[0179] Through fitting of historical data:

[0180] a represents the deviation amount caused by the position increasing by 1 mm per unit wear coefficient, and the calculation formula is:

[0181]

[0182] b reflects the basic wear offset independent of the position, such as the guide rail installation error, and the calculation formula is:

[0183]

[0184] Where:

[0185] P DTC,k : represents the P calculated in the kth group of data DTC ;

[0186] ΔP real,k : represents the difference between the true position measured by the micrometer and the encoder position in the kth group of data;

[0187] W guide,k : represents the W calculated in the kth group of data guide .

[0188] The real-time position P of the mold clamping mechanism encoderOn the basis of combining the dynamic temperature compensation amount and the wear compensation amount, the true position of the mold clamping mechanism is obtained, including:

[0189] P target =P DTC +ΔP WCM

[0190] Where: P target is the corrected true position, with the unit of mm. Fusing the real-time encoder position, the dynamic temperature compensation amount and the wear compensation amount to calculate the true position effectively solves the positioning deviation caused by the superposition of multi-source errors; in this step, the three are aggregated according to the weighted or superposition model to obtain the corrected true position; this fusion algorithm is implemented at the software level without additional sensors or machining parts; the fused true position not only eliminates the temperature drift error and wear offset, but also filters out occasional measurement noise; performing a zeroing movement on this basis can ensure that the movement end point is highly consistent with the mechanical origin; this method can be implemented by firmware or PLC upgrade on existing injection molding machines, with the advantages of rapid deployment and upgrade iteration; accurate true position feedback can also provide reliable data for the upper-layer MES or CPS system, promoting intelligent manufacturing.

[0191] Filtering the signal of the monitored zero-point sensor to retain the true zero-point sensor signal, including:

[0192] Collect the signal of the zero-point sensor and record it as S raw ;

[0193] Perform anti-interference filtering on S raw to obtain the filtered signal S filered :

[0194] S filtered =IDWT(SoftThresholding(DWT(S raw ),λ))

[0195] Where:

[0196] DWT: Discrete Wavelet Transform, using the db4 wavelet basis;

[0197] SoftThresholding: Soft threshold processing function;

[0198] IDWT: Inverse Discrete Wavelet Transform, reconstructing the filtered signal;

[0199] λ: Dynamic threshold, adaptively adjusted according to the interference frequency, specifically:

[0200]

[0201] Where:

[0202] λ0: The basic threshold value, which is the reference of the sensor noise level; the reference of the background noise amplitude of the sensor when there is no effective signal, used to distinguish the real signal from the noise; when the mode-locking mechanism is stationary (no movement, no load), continuously collect the output signal of the sensor and calculate the root mean square value of the signal as the basic threshold value;

[0203] f: The current signal frequency component;

[0204] f noise : The main interference frequency; when the device is operating normally, collect the sensor signal, perform a fast Fourier transform on the signal, identify the high-frequency peaks with significant amplitudes, and record the frequency corresponding to the peak as the main interference frequency;

[0205] σ: The frequency bandwidth coefficient; collect multiple groups of signals containing interference, analyze the distribution range of the interference frequency through a fast Fourier transform, and calculate the standard deviation of the frequency distribution as the frequency bandwidth coefficient;

[0206] If within three consecutive cycles, S filtered > λ, it is determined that the real mechanical zero point is detected. Wavelet anti-interference filtering of the zero point sensor signal and the use of an adaptive dynamic threshold to retain the real signal significantly improve the signal reliability; in this step, first perform wavelet decomposition on the collected discrete signal, use soft threshold processing to remove high-frequency interference components; then reconstruct the filtered signal through inverse transformation; the dynamic threshold is adaptively adjusted according to the main interference frequency and the frequency band width coefficient in the real-time signal spectrum, making the boundary between noise and signal more accurate; it is only when the filtered signal continuously meets the conditions within three consecutive cycles that the real mechanical zero point can be determined; this multi-level filtering and threshold strategy has stronger robustness and anti-interference ability compared to the fixed threshold scheme; it can also stably detect the true zero point signal in a strong electromagnetic noise and high-frequency vibration environment in the workshop, avoiding false touches and missed touches.

[0207] Moving the drive mode-locking mechanism until it returns to the mechanical zero point to complete the zeroing of the mode-locking mechanism includes:

[0208] Set the maximum allowable zeroing error value, denoted as ∈;

[0209] In each cycle, perform position error calculation:

[0210] μ = P target - P encoder

[0211] Where:

[0212] P target : The real position of the mode-locking mechanism obtained by combining the dynamic temperature compensation amount and the wear compensation amount within the cycle;

[0213] P encoder: It is the position of the mold clamping mechanism displayed by the encoder within a cycle;

[0214] When the true mechanical zero point is detected and |μ| < v, the mold clamping mechanism is controlled to stop, and the zeroing of the mold clamping mechanism is completed. Driving the mold clamping mechanism to move towards the mechanical zero point according to the true position and automatically stopping in combination with the maximum allowable error can achieve precise zeroing; in this step, the error between the true position and the encoder position is calculated in real time for each cycle, and the error is compared with the set maximum zeroing error value; when the true mechanical zero point is detected and the error is within the allowable range, the driver is automatically stopped; if the error exceeds the limit, the system will move in the reverse direction or make a fine adjustment at a reduced speed until the conditions are met; this closed-loop iterative strategy avoids excessive impact and control oscillation, ensuring stable convergence of mold clamping zeroing.

[0215] Embodiment 2, A system for implementing the mold clamping zeroing control method of the injection molding machine includes:

[0216] Encoder module: Displays the position of the mold clamping mechanism on the mold clamping mechanism;

[0217] Temperature detection module: Used to collect the temperatures of the tie rods, templates, and guide rails;

[0218] Zero point sensor module: Used to detect and identify the mechanical zero point;

[0219] Main controller module: Executes the multi-modal compensation algorithm;

[0220] Signal filtering module: Filters the signals collected by the zero point sensor module;

[0221] Clock synchronization module: Provides a unified time reference for the system to ensure the synchronization of data acquisition and control cycles;

[0222] Drive module: Drives the mold clamping mechanism to move.

[0223] Data storage module: Records historical data for regression calculation of parameters.

[0224] Embodiment 3, For the mold clamping zeroing control method of the injection molding machine, assume that a mold clamping zeroing control is performed on an injection molding machine of model HT-850T;

[0225] It is known that: The environmental temperature fluctuation is 25°C - 38°C;

[0226] The cumulative number of cycles of the guide rail X = 1.2×10 6 ;

[0227] The initial length of the tie rod L1 = 2000mm;

[0228] The initial length of the template L2 = 800mm;

[0229] The initial length of the guide rail L3 = 1500mm;

[0230] The linear expansion coefficient α1 of the pull rod is 12×10 -6 °C -1 ;

[0231] The linear expansion coefficient α2 of the template is 16×10 -6 °C -1 ;

[0232] The linear expansion coefficient α3 of the guide rail is 10×10 -6 °C -1 ;

[0233] The reference temperature T ref = 20°C;

[0234] The dynamic compensation coefficient β is 0.1 mm·s / °C;

[0235] The temperature-position factor γ is 1.05;

[0236] The wear rate s is 1×10 -6 times -1 ;

[0237] The linear coefficients a = 0.002 and b = 0.1;

[0238] The allowable error ∈ is 0.02 mm;

[0239] The collected data includes: the encoder position P encoder = 1500 mm;

[0240] The temperatures of the pull rod, template, and guide rail are T1 = 38°C, T2 = 35°C, and T3 = 36°C respectively;

[0241] The original sensor signal S raw = 120 mV, which includes 1 kHz interference;

[0242] Perform dynamic temperature compensation:

[0243] The expansion amount of the pull rod: ΔL1 = 12×10 -6 ×2000×(38 - 20) = 0.432 mm;

[0244] The expansion amount of the template: ΔL2 = 16×10 -6 ×800×(35 - 20) = 0.192 mm;

[0245] The expansion amount of the guide rail: ΔL3 = 10×10 -6 ×1500×(36 - 20) = 0.240 mm;

[0246] The dynamic temperature change compensation term:

[0247] Total expansion: ΔL = ΔL1 + ΔL2 + ΔL3 + ΔL dynamic = 10.864 mm;

[0248] Position after temperature compensation:

[0249] P DTC = P encoder - ΔL·γ = 1500 - 10.864×1.05 = 1488.593 mm

[0250] Calculate wear coefficient: W guide = s·X = 1×10 -6 ×1.2×10 6 = 1.2;

[0251] Wear compensation: ΔP WCM = 1.2(0.002×1488.593 + 0.1) = 3.692 mm

[0252] Obtain the corrected target position:

[0253] P target = 1488.593 + 3.692 = 1492.285 mm

[0254] Perform anti-interference filtering:

[0255] Where: λ0 = 50 mV; f = 1000 Hz; f noise = 1000 Hz; σ = 100 Hz;

[0256] Then the dynamic threshold

[0257] Decompose the original signal, S raw = 120 mV, filter out high-frequency noise (1 kHz interference);

[0258] Reconstructed signal S filtered = 75 mV;

[0259] Assume that the signal values for 3 consecutive cycles are: 75 mV, 78 mV, 80 mV;

[0260] Then it is determined that the true mechanical zero point is detected;

[0261] Perform position error calculation:

[0262] μ = P target - P encoder = 1492.285 - 1500 = -7.715 mm

[0263] The error is negative, control the mode-locking mechanism to move in the reverse direction;

[0264] The moved position is P encoder = 1493.285 mm;

[0265] The remaining error is μ = 1492.285 - 1493.285 = -1 mm;

[0266] Since the error is negative, control the reverse movement of the mode-locking mechanism;

[0267] The moved position is P encoder = 1492.285 mm;

[0268] The final error is μ = 1492.285 - 1492.285 = 0 mm;

[0269] If the true mechanical zero point is detected at this time, the mode-locking zeroing control is completed.

[0270] It should be noted that in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variation thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device.

[0271] The above are only the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the technical principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. Injection molding machine clamping zeroing control method, characterized in that, Including: Taking 10 ms as a cycle, and performing mode-locking zero calibration control within each cycle, specifically: Obtain the real-time position of the mode-locking mechanism through the encoder, denoted as P encoder ; Collecting the temperatures of the tie bar, the platen, and the guide rail through a temperature detection module, and respectively denoting them as T1, T2, and T3; Calculating the thermal expansion amounts of each component according to the temperatures of the tie bar, the platen, and the guide rail, and taking the obtained thermal expansion amounts as dynamic temperature compensation amounts; Calculating the motion trajectory deviation caused by mechanical wear according to the number of cycles of the guide rail of the current injection molding machine, and taking it as the wear compensation amount; Based on the real-time position P of the mold clamping mechanism encodr and combining the dynamic temperature compensation amount and the wear compensation amount, the real position of the mold clamping mechanism is obtained; Controlling the die locking mechanism to move towards the mechanical zero point according to the actual position of the die locking mechanism; Monitoring the signal state of the zero point sensor during the process of the die locking mechanism approaching the mechanical zero point; Filtering the monitored signal of the zero point sensor to retain the true zero point sensor signal; Driving the die locking mechanism to move until it returns to the mechanical zero point, and completing the zeroing of the die locking mechanism.

2. The injection molding machine clamping return-to-zero control method according to claim 1, characterized in that, The calculating the thermal expansion amounts of each component according to the temperatures of the tie bar, the platen, and the guide rail, and taking the obtained thermal expansion amounts as dynamic temperature compensation amounts includes: Calculating the thermal expansion amounts of the tie bar, the platen, and the guide rail: Where: α i : The linear expansion coefficients of the tie rod, formwork, and guide rail are obtained according to the materials of each component. L i : Initial lengths of the tie rod, formwork, and guide rail, in mm; T ref : Reference temperature, set to 20 °C; β: dynamic compensation coefficient, calibrated through a thermal shock experiment; T i : Collect the temperatures of the tie rod, formwork, and guide rail through a temperature sensor, which are T1, T2, and T3 respectively; The temperature change rate is obtained by calculating the temperature difference between adjacent cycles: Where: T i (t): the temperature of this cycle; T i (t - 1): the temperature of the previous cycle; The position data after removing the thermal expansion amount is denoted as P DTC ; P DTC = P encoder -ΔL·γ Where: γ: representing the temperature-position factor, obtained by calculating through historical data.

3. The injection molding machine clamping return-to-zero control method according to claim 2, characterized in that The calibrating through a thermal shock experiment is specifically: Performing M experiments on the die locking mechanism, and applying temperature steps at different rates in each experiment; In each experiment, record the temperature change rate and the actual expansion amount ΔL real,k , fit β by the least squares method, take the derivative of the error function and set the derivative to zero: Where: M: representing the total number of samples of the experimental data, that is, the number of thermal shock experimental data points used to calibrate β; ΔL real,k : Measured thermal expansion during the k-th experiment, unit: mm; ΔT k : The difference between the temperature collected and the reference temperature in the k-th experiment; The temperature change rate of the k-th experiment.

4. The injection molding machine mold clamping zeroing control method according to claim 2, characterized in that The obtaining by calculating through historical data includes: Collecting V groups of temperature, encoder position, and true position data of manual calibration under the no-load state of the equipment; Calculate ΔL for the k-th group of data k ; Calculating the temperature-position factor as: Where: P encoder,k : The encoder position measured in the k-th set of data, in mm; P real,k : The true position of manual calibration in the k-th group of data, with the unit of mm; ΔL k : The thermal expansion amount calculated from the k-th group of data, with the unit of mm; V: the number of groups of collected data.

5. The injection molding machine clamping return-to-zero control method according to claim 1, characterized in that The calculating the motion trajectory deviation caused by mechanical wear according to the number of cycles of the guide rail of the current injection molding machine, and taking it as the wear compensation amount includes: Denote the wear compensation amount as ΔP WCM , then ΔP WCM = W guide ·(a·P DTC + b) Where: W guide : Wear coefficient, obtained by calculating the number of cyclic operations; a, b: linear coefficients, obtained by regression calculation through historical data; The obtaining by calculating through the number of cycles is specifically: W guide = s·X X: the number of cycles of the guide rail of the current injection molding machine; s: wear coefficient, obtained according to the statistical result of the industry average wear rate of the guide rail of the injection molding machine.

6. The injection molding machine clamping return-to-zero control method according to claim 5, characterized in that, The linear coefficients, obtained by regression calculation through historical data, are specifically: Obtain Z groups of historical data, each group of data including the position P after temperature compensation DTC,k , the actual wear deviation ΔP real,k , the wear coefficient W guide,k ; Fitting through historical data: a represents the deviation amount caused by the position increasing by 1 mm per unit wear coefficient, and the calculation formula is: b reflects the basic wear offset amount independent of the position, and the calculation formula is: Where: P DTC,k : It represents the P calculated from the k-th group of data DTC ; ΔP real,k : represents the difference between the true position measured by the dial indicator and the encoder position in the k-th group of data; W guide,k : It represents the W calculated from the k-th group of data guide .

7. The injection molding machine clamping zeroing control method according to claim 1, characterized in that, Based on the real-time position P of the mold clamping mechanism encoder Based on this, by combining the dynamic temperature compensation amount and the wear compensation amount, the real position of the mold clamping mechanism is obtained, including: P target = P DTC + ΔP WCM Where: P target is the corrected true position in mm.

8. The injection molding machine clamping zeroing control method according to claim 1, characterized in that, The filtering the monitored signal of the zero point sensor to retain the true zero point sensor signal includes: Collect the signal of the zero-point sensor and denote it as S raw ; Perform anti-interference filtering on S raw to obtain the filtered signal S filtered : S filtered = IDWT(SoftThresholding(DWT(S raw ), λ)) Where: DWT: discrete wavelet transform, using the db4 wavelet basis; SoftThresholding: soft threshold processing function; IDWT: inverse discrete wavelet transform, reconstructing the filtered signal; λ: dynamic threshold, adaptively adjusted according to the interference frequency, specifically: Where: λ0: basic threshold, the reference of the sensor noise level; f: the current signal frequency component; f noise : Main interference frequency; σ: frequency bandwidth coefficient; If within three consecutive cycles, S filtered > λ, it is determined that the true mechanical zero point has been detected.

9. The injection molding machine clamping return-to-zero control method according to claim 1, wherein The driving the die locking mechanism to move until it returns to the mechanical zero point, and completing the zeroing of the die locking mechanism includes: Setting the maximum allowable zeroing error value, denoted as ∈; In each cycle, the position error is calculated: μ = P target -P encoder Where: P target : The actual position of the mode-locking mechanism obtained by combining the dynamic temperature compensation amount and the wear compensation amount within the period; P encoder : The position of the mode-locking mechanism displayed by the encoder within the period; When the true mechanical zero point is detected and |μ| < ∈, the control of the mold clamping mechanism stops, and the mold clamping mechanism is reset to zero.

10. A system adopting the injection molding machine clamping die zeroing control method described in claim 1, characterized in that, Including: Encoder module: Displays the position of the mold clamping mechanism on the mold clamping mechanism; Temperature detection module: Used to collect the temperatures of the tie rod, template, and guide rail; Zero point sensor module: Used to detect and identify the mechanical zero point; Main controller module: Executes the multi-modal compensation algorithm; Signal filtering module: Filters the signals collected by the zero point sensor module; Clock synchronization module: Provides a unified time reference for the system to ensure the synchronization of data acquisition and control cycles; Drive module: Drives the movement of the mold clamping mechanism. Data storage module: Records historical data for regression calculation of parameters.

Images