Helical angle error online detection system of gear hobbing machine executing mechanism

Abstract

The invention provides a gear hobbing machine execution mechanism helical angle error on-line detection system comprising a first angular displacement sensor used for collecting workbench angular displacement data; the second angular displacement sensor is arranged on a gear hobbing machine cutter bar shaft and used for collecting cutter bar angular displacement data; the linear displacement sensor is arranged on the gear hobbing machine tool rest and used for collecting linear displacement data of the tool rest in the vertical direction; the data acquisition card is used for processing the angular displacement data of the workbench, the angular displacement data of the cutter bar and the linear displacement data of the cutter rest in the vertical direction according to the current main time base, and the current main time base is dynamically switched according to the channel stability score of each channel; and the processor is connected with the data acquisition card and is used for determining the helical angle error of the helical gear according to the processed angular displacement data of the workbench, the angular displacement data of the cutter bar and the linear displacement data of the cutter rest in the vertical direction. According to the invention, the detection precision of the helical angle error of the gear hobbing machine execution mechanism can be improved.

Description

Technical Field

[0001] This invention belongs to the field of mechanical precision testing technology, specifically relating to an online detection system for the helix angle error of a gear hobbing machine actuator. Background Technology

[0002] In high-end equipment manufacturing, the actuator of the gear hobbing machine is the source of precision in helical gear machining. The worktable rotates to achieve workpiece indexing, the tool holder shaft drives the hob to complete the cutting, and the tool holder moves in the vertical direction to form the helical lead. The synchronicity of the motion and the displacement accuracy of the three directly determine the machining quality of the helix angle of the helical gear.

[0003] During the factory acceptance testing of gear hobbing machines, performance verification is typically performed by trial-cutting a standard helical gear. This involves machining a standard helical gear test piece under rated process parameters and sending it to a precision gear measurement center for testing. By analyzing key indicators in the test report, such as helical shape deviation, helical tilt deviation, and total tooth direction deviation, the machine tool's actuator is evaluated to assess whether it can accurately achieve the helical motion required for helical gear machining, thereby determining whether it meets the design accuracy requirements. The main limitation of this method is that it cannot identify motion anomalies in real time during machining.

[0004] To overcome the above shortcomings, existing literature has proposed a transmission error detection approach centered on "synchronous displacement comparison between output and input": simultaneously acquiring the displacement of the workpiece and hobbing cutter channels while the equipment is running, and evaluating the positional deviation through the displacement difference; when this approach is combined with the multi-axis motion (generating and differential) relationship of the gear hobbing machine, it is expected to reduce the sampling error caused by time base jitter and has the potential for mechanization. However, existing solutions mostly use a fixed time interval as the sampling base, without considering the stability differences of the sensor channels of each actuator under the on-site working conditions of the gear hobbing machine. A fixed time base not only cannot adapt to the dynamic characteristics of different channels, but also amplifies the sampling alignment deviation due to the instantaneous speed jitter of the spindle, and the accumulated error directly destroys the coupling relationship of the multi-axis motion; therefore, existing implementations still have the problem of insufficient accuracy in detecting the helix angle error of the gear hobbing machine actuator, resulting in low gear machining accuracy. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide an online detection system for the helix angle error of the gear hobbing machine actuator, so as to meet the need to improve the detection accuracy of the helix angle error of the gear hobbing machine actuator, thereby improving the working accuracy of the gear hobbing machine.

[0006] To achieve the above objectives, the present invention provides the following technical solution: According to a first aspect, the present invention provides an online detection system for the helix angle error of a gear hobbing machine actuator, comprising: a first angular displacement sensor, disposed on the gear hobbing machine worktable, for collecting worktable angular displacement data; a second angular displacement sensor, disposed on the gear hobbing machine tool holder shaft, for collecting tool holder angular displacement data; a linear displacement sensor, disposed on the gear hobbing machine tool holder, for collecting tool holder vertical linear displacement data; a data acquisition card, connected to the first angular displacement sensor, the second angular displacement sensor, and the linear displacement sensor respectively, for processing the worktable angular displacement data, tool holder angular displacement data, and tool holder vertical linear displacement data according to a current master time base, wherein the current master time base is dynamically switched according to the channel stability score of each channel; and a processor, connected to the data acquisition card, for determining the helix angle error of the helical gear based on the processed worktable angular displacement data, tool holder angular displacement data, and tool holder vertical linear displacement data.

[0007] This invention provides an online detection system for the helix angle error of a gear hobbing machine actuator. Three-channel synchronous displacement sampling and real-time FPGA processing are embedded in the machine tool workflow, enabling rapid on-machine detection. This reduces additional errors caused by off-machine handling and secondary clamping, shortening the production cycle. Furthermore, a dynamic switching of the master time base is employed, identifying the channel with the highest stability score and switching it to the master time base. This ensures that the data from the other two channels are aligned with the optimal benchmark, avoiding the cascading effects of fluctuations in a single channel. The data acquisition card dynamically switches the master time base based on the stability scores of each sensor channel. This adapts to the dynamic characteristics differences of various actuators under the on-site working conditions of the gear hobbing machine and effectively suppresses sampling alignment deviations caused by factors such as instantaneous spindle speed fluctuations and vibrations. It avoids error accumulation under a fixed time base, ensuring the synchronization and accuracy of multi-axis displacement data, significantly improving the accuracy of helix angle error detection for gear hobbing machine actuators, and benefiting subsequent gear machining by reducing gear machining errors.

[0008] Other advantages, objectives, and features of the invention will be set forth in the following description and will be apparent to those skilled in the art in some respects, or may be learned by practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0009] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the following figures are provided for illustration: Figure 1 This is a specific example diagram of an online detection system for the helix angle error of a gear hobbing machine actuator according to the present invention; Figure 2 This is a schematic diagram of the three-axis motion state of the gear hobbing machine in this invention. Detailed Implementation

[0010] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0011] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can also refer to the internal connection of two components; and they can refer to a wireless connection or a wired connection. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0012] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0013] This invention provides an online detection system for the helix angle error of a gear hobbing machine actuator, such as... Figure 1 As shown, it includes: The first angular displacement sensor 101 is installed on the gear hobbing machine table to collect the angular displacement data of the table. The second angular displacement sensor 102 is installed on the tool bar shaft of the gear hobbing machine to collect tool bar angular displacement data; Linear displacement sensor 103 is installed on the tool holder of the gear hobbing machine to collect linear displacement data in the vertical direction of the tool holder. The data acquisition card 104 is connected to the first angular displacement sensor 101, the second angular displacement sensor 102 and the linear displacement sensor 103 respectively, and is used to process the table angular displacement data, tool holder angular displacement data and tool post vertical linear displacement data according to the current master time base. The current master time base is dynamically switched according to the channel stability score of each channel. The processor 105 is connected to the data acquisition card 104 and is used to determine the helical gear helix angle error based on the table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data after the current master time base processing.

[0014] For example, such as Figure 2 The diagram shown is a schematic of the three-axis motion of a gear hobbing machine, where the table angular displacement data corresponds to... The tool holder angular displacement data is The vertical linear displacement data of the tool holder is LThe first angular displacement sensor 101 is installed on the gear hobbing machine's worktable to collect the worktable's angular displacement data. The second angular displacement sensor 102 is installed on the tool holder shaft of the gear hobbing machine to collect the angular displacement data of the tool holder. Linear displacement sensor 103 is installed on the tool holder of the gear hobbing machine to collect linear displacement data in the vertical direction of the tool holder. L This forms a three-variable measurement chain that is consistent with the relationship between the gear hobbing generation and the differential.

[0015] To convert the analog signals acquired by the sensors into digital signals via a circuit shaping module, the data from the three sensor channels is conditioned (filtered, shaped, isolated, and level matched) before being sent to the data acquisition card 104. The data acquisition card 104 can be constructed using an FPGA. It processes the table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data based on the current master time base, which dynamically switches according to the channel stability score of each channel. Next, it interacts with the processor 105 through a communication interface with frame sequence numbers and CRC checks. The link side is equipped with an asynchronous FIFO and a retransmission buffer to ensure data integrity and timing continuity. The processor 105 can be a host computer, which can be a device or module with computing capabilities, such as a computer terminal, used to determine the helical gear helix angle error based on the table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data processed by the current master time base. In other words, this embodiment adopts an integrated architecture of "three-channel synchronous displacement sampling + FPGA real-time alignment and compensation + host computer robust calculation".

[0016] During operation, after the system powers on and completes its self-test, it first performs self-calibration under no-load conditions, and then performs self-calibration in the spur gear section ( Under certain conditions, refined self-calibration is performed, and initial values ​​such as transmission ratio and phase zero point are estimated and written. During the measurement phase, the trigger edge of the "main time base channel" is used as a reference. The FPGA opens an equal-width time window before and / or after triggering, and uses a high-frequency clock to count and interpolate to the trigger time to obtain the equivalent displacement of the three channels at the same reference time, achieving sub-cycle-level synchronous alignment. It should be noted that dynamic switching of the main time base is introduced in this embodiment to improve the reliability of the entire error detection system and enable it to quickly adapt to changes in the scene, thereby improving adaptability.

[0017] In this embodiment, the primary time base switches based on changes in channel stability scores. The system includes three channels: a data transmission channel between the first angular displacement sensor 101 and the data acquisition card 104; a data transmission channel between the second angular displacement sensor 102 and the data acquisition card 104; and a data transmission channel between the linear displacement sensor 103 and the data acquisition card 104. The primary time base switches according to the channel stability scores, which are determined within a sliding window. The stability scores include trigger interval variance, root mean square of the dual-window interpolation alignment residuals, effective signal-to-noise ratio, and bit loss rate per unit time.

[0018] Specifically, the current master time base is dynamically switched based on the channel stability scores of each channel, including: acquiring the historical trigger edge signal time interval of the three channels, the historical channel signals of the three channels, the equivalent values ​​of each historical channel signal at the trigger edge signal time of the historical master time base, and the actual sampled values ​​within the equivalent value calculation window; determining the historical trigger interval variance of each channel based on the historical trigger edge signal time interval of the three channels; determining the root mean square of the interpolation alignment residual of each channel based on the equivalent values ​​of each historical channel signal at the trigger edge signal time of the historical master time base and the actual sampled values ​​within the equivalent value calculation window; determining the signal-to-noise ratio and bit loss rate per unit time of each channel based on the historical channel signals of the three channels; evaluating the channel stability score based on the trigger interval variance, root mean square of the interpolation alignment residual, signal-to-noise ratio, and bit loss rate per unit time of each channel; and determining the current master time base based on the channel stability score.

[0019] Since all sensors in this embodiment need to continuously monitor the temporal variation of the corresponding physical quantities, equal time intervals between trigger edge signals maximize data validity. Therefore, the trigger interval is used as one of the parameters for stability evaluation. The historical trigger edge signal time interval can represent the trigger edge signal time interval before stability scoring. The historical trigger edge signal time intervals of the three channels can be obtained by retrieving a pre-stored file in the external memory. This file can be obtained through real-time monitoring by software. That is, when a trigger edge is detected, the clock of the current operating system is read, a timestamp is generated, a timestamp file of the trigger edge is obtained, and it is stored in the memory. Similarly, the historical channel signals of the three channels, the equivalent values ​​of each historical channel signal at the historical master time base trigger edge signal time, and the actual sampled values ​​can also be obtained by reading the external memory. This embodiment will not elaborate further, but it should be noted that since the historical master time base trigger edge signal time may not match the sampling time, it is impossible to obtain the actual sampled value at the historical master time base trigger edge signal time. Therefore, in this embodiment, the closest actual sampled value within the equivalent value calculation window is collected as the actual sampled value. That is, the query range of the actual sampled value is expanded. The equivalent value can be calculated by dual-window interpolation, that is, one window is taken before and after the historical master time base trigger edge signal time, and the two windows before and after realize interpolation.

[0020] After obtaining the historical trigger edge signal time intervals of the three channels, the historical channel signals of the three channels, the equivalent values ​​of each historical channel signal at the historical master time base trigger edge signal time, and the actual sampled values ​​within the equivalent value calculation window, the historical trigger interval variance, the root mean square of the interpolation alignment residuals of each channel, the signal-to-noise ratio, and the bit loss rate per unit time of each channel are determined. Then, the channel stability score is determined based on the following formula. : ; in, , , , These are the historical trigger interval variance of the channel, the root mean square of the interpolation alignment residual of the channel, the signal-to-noise ratio of the channel, and the bit loss rate per unit time, respectively. , , , The above formula, with corresponding weights, normalizes each indicator into a score that follows the same direction. , , , Then, based on the weights, a single scoring weight and a normalized scale can be configured, and for... A slight time smoothing is applied to suppress transient fluctuations. Channel stability is used as one of the factors for switching the primary time base; the more stable the channel, the more likely it is to be used as the primary time base again.

[0021] For example: during initialization, w = [ , , , =[0.16, 0.25, 0.44, 0.15], and then recalculate the weights every 3 rotations of the worktable. The collected data is processed as follows to update the weights: (1) Data standardization processing, using indicator formulas, such as: ,in, For the first i The evaluation object is in the first j Standardized values ​​on a positive indicator For the first i The evaluation object is in the first j The original value on a positive indicator For the first j The maximum value of a positive indicator among all evaluation objects. For the first jThe minimum value of a positive indicator among all evaluation objects; it should be noted that the above formula can be used for positive indicators (the larger the value, the better), while the following formula can be used for negative indicators (the smaller the value, the better): ; (2) Calculate the probability matrix: ; (3) Calculate the entropy value: ; (4) Calculate the weights: .

[0022] Once the current master time base is determined, the processor 105 aligns the table angular displacement data, tool bar angular displacement data, and tool post vertical linear displacement data obtained from the three sensor channels based on the master time base. The processor 105 then determines the helical gear helix angle error based on the following formula and the aligned table angular displacement data, tool bar angular displacement data, and tool post vertical linear displacement data. .

[0023] ; Where z represents the number of teeth, m is the normal module, and K is the number of worm threads. The theoretical helix angle, This represents the angular displacement of the tool holder. This represents the angular displacement of the worktable during rotation. L The ± sign in the formula is determined based on the machining method, representing the linear displacement of the tool holder. The above formula is existing technology; the derivation process will not be elaborated upon.

[0024] This invention provides an online detection system for the helix angle error of a gear hobbing machine actuator. Three-channel synchronous displacement sampling and real-time FPGA processing are embedded in the machine tool workflow, enabling rapid on-machine detection. This reduces additional errors caused by off-machine handling and secondary clamping, shortening the production cycle. Furthermore, a dynamic switching of the master time base is employed, identifying the channel with the highest stability score and switching it to the master time base. This ensures that the data from the other two channels are aligned with the optimal benchmark, avoiding the cascading effects of fluctuations in a single channel. The data acquisition card dynamically switches the master time base based on the stability scores of each sensor channel. This adapts to the dynamic characteristics differences of various actuators under the on-site working conditions of the gear hobbing machine and effectively suppresses sampling alignment deviations caused by factors such as instantaneous spindle speed fluctuations and vibrations. It avoids error accumulation under a fixed time base, ensuring the synchronization and accuracy of multi-axis displacement data, significantly improving the accuracy of helix angle error detection for gear hobbing machine actuators, and benefiting subsequent gear machining by reducing gear machining errors.

[0025] As an optional implementation, the table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data are processed based on the current master time base. This includes: upon receiving a trigger edge signal from the current master time base, acquiring a first time window and a second time window with the same time width before and after the trigger edge signal; counting the three channel signals within the first and second time windows according to the target clock to obtain first count data for each channel signal in the first time window and second count data in the second time window, where the target clock frequency is higher than the frequency of the three channel signals, and the three channels are the first angular displacement sensor channel, the second angular displacement sensor channel, and the linear displacement sensor channel; and using an interpolation algorithm based on the first and second count data to determine the equivalent value of each channel signal at the trigger edge signal time, which is then used as the processed table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data.

[0026] For example, firstly, an FPGA is used as the core control unit, paired with a high-frequency target clock generated by a PLL from the main time base clock, such as 100MHz, with a counting resolution of 10ns. The frequency of the high-frequency target clock is greater than the highest frequency of the three sensors. Simultaneously, the signals from the three sensors are shaped by a Schmitt trigger and then connected to the FPGA data acquisition card. When the FPGA data acquisition card detects the trigger edge of the main time base... First, latch the high-frequency clock count value corresponding to the trigger moment, and immediately generate two time windows of equal width before and after. The width of the time window can be set to twice the sensor period to ensure the capture of the complete pulse. The first time window before ~ ]and The second time window after [ ~ ].

[0027] Subsequently, independent counting units were assigned to the three channels to count the number of sensor pulses within the first time window. Record the high-frequency count value of the last pulse edge and count the number of pulses within the second time window. The high-frequency count value of the first pulse edge is recorded to obtain the first and second count data for each channel. Next, the count data is calculated by interpolating the last edge before triggering with... Time difference, first edge after triggering and The time difference, combined with the sensor pulse equivalent, will , The values ​​are converted to corresponding displacement values, and then linear interpolation is used to calculate the equivalent value of the trigger time. Finally, the three interpolation results are buffered by FIFO and output in the same clock cycle to obtain time-synchronized table angular displacement, tool holder angular displacement, and tool post vertical linear displacement data.

[0028] Specifically, based on the high-frequency clock count captured by the FPGA, the last signal edge before triggering is calculated. and Time difference and the first signal edge after triggering. and Time difference To determine exist - The positional proportion within the time interval; then, combined with the sensor pulse equivalent, the counting data within the first and second time windows are analyzed. , Convert to actual displacement value , ,for example, Multiply by the pulse equivalent to obtain And specify the displacement step size corresponding to a single pulse. Subsequently, linear interpolation is applied to the pulse signal, and a portion of the pulse displacement is allocated according to time proportions to achieve sub-pulse level accuracy, ensuring that the output table angular displacement, tool holder angular displacement, and tool post vertical linear displacement data are strictly aligned in time. This achieves sub-period-level synchronization. Compared to traditional single-point timestamps, the "dual-window" interpolation can evenly trigger jitter and improve robustness to speed fluctuations. Simultaneously, the root mean square of the interpolation residual *e* is calculated within the same window and incorporated as a channel quality factor in subsequent scoring. Furthermore, it allows for adaptive fine-tuning of the window width within a defined range to balance alignment accuracy and noise suppression. To suppress the cumulative offset caused by slow-speed waves and mechanical jitter, the system estimates the alignment residual within adjacent periods and performs feedforward correction, forming a continuous correction chain for "phase drift compensation."

[0029] This invention provides an online detection system for the helix angle error of a gear hobbing machine actuator. By using front and rear dual-window interpolation combined with phase drift compensation, sub-cycle-level time alignment is achieved, making the helix angle curve smoother and more consistent, and less sensitive to instantaneous fluctuations and trigger jitter of the spindle.

[0030] As an optional implementation, determining the current primary time base based on channel stability scores includes: determining candidate channels based on channel stability scores; when a candidate channel is inconsistent with the primary time base channel after the last switch and the time since the last primary time base switch exceeds a preset time length, determining whether the difference between the stability score of the candidate channel in the target time period and the stability score of the primary time base channel after the last switch is continuously greater than a first threshold, and / or whether the difference between the stability score of the candidate channel and the stability score of the primary time base channel after the last switch is greater than a second threshold, wherein the second threshold is less than the first threshold; when the difference between the stability score of the candidate channel in the target time period and the stability score of the primary time base channel after the last switch is continuously greater than the first threshold, and / or whether the difference between the stability score of the candidate channel and the stability score of the primary time base channel after the last switch is greater than the second threshold, then switching the time base of the candidate channel to the primary time base.

[0031] For example, after the candidate channel pool is determined, to prevent frequent switching of the primary time base in a short period of time from causing system timing disorder, it is necessary to first determine whether the hysteresis criterion is met. Specifically, the channel and switching time of the last primary time base switch are extracted, and it is determined whether the last primary time base was the primary time base of the candidate channel, and whether the difference between the current time and the last switching time is greater than a preset time length. If the candidate channel is the current primary time base channel, no switching is required, and the process terminates. If the difference between the current time and the last switching time is less than or equal to the preset time length, the next judgment is not performed until the difference is greater than the preset time length. That is, the next judgment is performed only when both judgment conditions are met simultaneously: the candidate channel is not the previous primary time base channel, and the difference between the current time and the last switching time is greater than the preset time length.

[0032] The next crucial step is to verify the stability advantage of the candidate channel using two thresholds, and determine whether a switch should occur according to the following rules: 1) Candidate Channel j Satisfy within the target time period When a switch is triggered, the target time period can be any of three consecutive switch cycles. Assess the stability of candidate channel j. The stability score is the score for the primary time base channel determined during the last primary time base switch (i.e., the primary time base channel before the current switch). The first threshold; 2) If the current channel score recovers and meets the requirements during the observation period... ( Less than ), keep not switching, among which, The second threshold is set; after a switch is completed, no new switch decision is made within the set cooling time. This embodiment does not limit the cooling time, but those skilled in the art can determine it as needed.

[0033] In addition, during acceleration, emergency stop, or link anomaly phases, evaluation and switching are paused to maintain the current master time base, ensuring that judgments are based on comparable data. This invention provides an online detection system for the helix angle error of a gear hobbing machine actuator, which avoids frequent switching and improves stability through limiting conditions.

[0034] As an optional implementation, after the current master time base is dynamically switched based on the channel stability scores of each channel, this embodiment selects several trigger points on both sides of the switching point to form an overlapping area. The least squares method is used to estimate the ratio and offset between the old and new sequences, and a one-time correction is performed on the new master time base. If there is a small time offset, a short-window cross-correlation can be performed within the overlapping area to eliminate the phase difference. Specifically, a first target sequence before the switching point and a second target sequence after the switching point are selected, where the first and second target sequences are physically overlapping. The cross-correlation function between the first and second target sequences is calculated to determine the time offset between them. Based on the time offset, the second target sequence is aligned. A relationship model between the first target sequence and the adjusted second target sequence is determined using a fitting algorithm. Based on the relationship model, the channel signals after the switching point are corrected.

[0035] For example, the physical time of the preset primary time base switching is denoted as T. Sensor data from T-1s to T+0.5s is collected and denoted as X=[ , ,..., As the first target sequence, the sampling rate is f=1kHz, and each data point is accompanied by an absolute timestamp, such as... =T-1s, =T+0.5s. Collect sensor data from T-0.5s to T+1s, denoted as Y=[ , ,..., As the second target sequence, the sampling rate is the same as X, and it also includes an absolute timestamp. =T-0.5s, =T+1s. The physical time overlap interval between the two sequences is [T-0.5s, T+0.5s], ensuring that the sensor measures the same physical process within this interval.

[0036] Short window truncation extracts a short window of data from the overlapping region [T-0.5s, T+0.5s], and extracts the data from X corresponding to [T-0.1s, T+0.1s] as the first truncation sequence. The data corresponding to [T-0.1s, T+0.1s] in Y is extracted as the second extraction sequence. The cross-correlation function is calculated for the first and second truncated sequences using the following formula: in, M is the time offset, and M is the window length. Represents the first in the sequence Data.

[0037] Based on the above formula, find The maximum value corresponding to ,like This means that the new sequence lags behind the old sequence by 3 sampling points, corresponding to 3ms. This value is the time offset between the two sequences.

[0038] Next, based on Adjust the time axis of Y, specifically, when >0, shift the entire Y-axis forward. Each sampling point, i.e., before deletion One point, tail fill The nearest neighbor values ​​are used to obtain a new aligned sequence; when <0, shift the entire Y axis backward. |Sampling points, i.e., head padding| Find the nearest neighbor values ​​and remove the last one. | points. Finally, using X as the baseline, a linear fit is performed on X and the aligned second target sequence within the overlap region [T-0.5s, T+0.5s], and the proportion is calculated by least squares solution. and bias A linear relationship model is determined based on proportion and bias: Then, based on a linear relationship model, the channel signal after the switching point is corrected to obtain the corrected output data.

[0039] This invention provides an online detection system for the helix angle error of a gear hobbing machine actuator. When the master time base is switched, the measurement data may jump at the switching point due to differences in clock sources (such as frequency deviation and phase shift). By aligning the cross-correlation time in the overlapping area, the small phase difference can be eliminated. Then, the proportional and bias models are obtained by least squares fitting, and the new sequence is corrected in one go, so that the data before and after the switch are seamlessly connected in both numerical and time dimensions, avoiding the interference of the jump to subsequent analysis.

[0040] As an optional implementation, after correction, the output is weighted in the transition region according to monotonic weights from old to new, with the sum of the weights remaining at one. The weights gradually transition from one to zero, thereby avoiding step and phase misalignment. This transition region can be the same area as the overlap region. Therefore, the online detection system for helix angle error of a gear hobbing machine actuator proposed in this embodiment further includes: weighting a first target sequence according to a first weighted sequence, with the values ​​of the first weighted sequence arranged from high to low; and weighting a corrected second target sequence according to a second weighted sequence, with the values ​​of the second weighted sequence arranged from low to high. The sum of the weights of the first weighted sequence and the second weighted sequence at the same physical time is 1.

[0041] For example, the first weighted sequence generates the first weighted sequence. The values ​​must be arranged from highest to lowest, with a weight of 1 at the beginning of the overlapping region (T0-0.5s) and a weight of 0 at the end region (T0+0.5s). A linear decreasing method is used for generation. ; The second weighted sequence generates the second weighted sequence. The values ​​must be arranged from low to high, and must be consistent with... The weights sum to 1 at the same physical time. Generated using a linearly increasing method: ; Extracting subsequences from the overlapping region of the first target sequence according to... Weighted subsequences are extracted from the overlapping region of the corrected second target sequence. The weighted, overlapping data fusion method results in the final output data of the two weighted sequences being superimposed, which enables a smooth transition in this embodiment.

[0042] As an optional implementation, in order to sense the working conditions and perform environmental compensation, temperature sensors and triaxial accelerometers are installed in key parts. Therefore, an online detection system for the helix angle error of a gear hobbing machine actuator also includes: a temperature sensor for collecting the working environment temperature of the gear hobbing machine; and an accelerometer for collecting vibration data of the gear hobbing machine. Based on the processed table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data, the helical gear helix angle error is determined, including: correcting the processed table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data based on the working environment temperature and vibration data; and determining the helical gear helix angle error based on the corrected table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data processed based on the current master time base.

[0043] For example, the temperature channel is used to perform linear or quadratic corrections to the calibration factor and zero point, while the vibration channel uses bandpass energy or RMS to indicate impact disturbances, deweighting or eliminating affected samples, and adding feedforward corrections to the short-time phase when necessary. The sensor's calibration factor and zero point drift due to temperature. The temperature channel establishes a mathematical relationship between temperature, calibration factor, and zero point by acquiring ambient or sensor-specific temperature data in real time, and uses this relationship to correct the main channel's measurements in real time, offsetting systematic errors caused by temperature changes.

[0044] The vibration channel is specifically designed to monitor external shocks, vibrations, and other interference. It determines the presence of significant disturbances by using bandpass energy or RMS. If the disturbance is strong, it indicates that the main channel measurement samples at the corresponding time may be distorted due to interference; in this case, these samples are downweighted or directly removed. If the disturbance is short-lived and mainly affects the phase of the measurement signal, feedforward correction is added. Based on the disturbance characteristics of the vibration channel, the phase correction amount is calculated in advance and applied to the main channel data to actively offset the phase deviation caused by the disturbance. This embodiment, through the detection of vibration and temperature, ultimately corrects the error calculation, thereby improving the accuracy of error calculation.

[0045] As an optional implementation, an online detection system for the helix angle error of a gear hobbing machine actuator further includes: a memory for storing table angular displacement data, tool holder angular displacement data, tool post vertical linear displacement data, time offset, relational model, and helix angle error of the helical gear. A weighted fitting of the helix angle and tooth profile curve is performed within a sliding window, and RANSAC can be used to remove outliers, outputting evaluation indicators such as maximum value, minimum value, peak-to-peak value, RMS, and dominant frequency component. All original counts, calibration parameters, and compensation coefficients are archived along with the measurement results, facilitating both recalculation and traceability, and enabling the generation of batch process capability reports.

[0046] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made to it in form and detail without departing from the scope defined by the claims of the present invention.

Claims

1. An online detection system for the helix angle error of a gear hobbing machine actuator, characterized in that, include: The first angular displacement sensor is installed on the gear hobbing machine's worktable to collect the worktable's angular displacement data. The second angular displacement sensor is installed on the tool holder shaft of the gear hobbing machine to collect tool holder angular displacement data; A linear displacement sensor is installed on the tool holder of the gear hobbing machine to collect linear displacement data in the vertical direction of the tool holder; The data acquisition card is connected to the first angular displacement sensor, the second angular displacement sensor, and the linear displacement sensor respectively. It is used to process the table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data according to the current master time base. The current master time base is dynamically switched according to the channel stability score of each channel. The processor, connected to the data acquisition card, is used to determine the helical gear helix angle error based on the processed table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data.

2. The online detection system for helix angle error of a gear hobbing machine actuator according to claim 1, characterized in that, Based on the current master time base processing of table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data, including: When the trigger edge signal of the current master time base is received, the three channels acquire a first time window and a second time window with the same time width before and after the trigger edge signal time, respectively. The three channel signals in the first and second time windows are counted separately according to the target clock to obtain the first count data of each channel signal in the first time window and the second count data of the second time window. The target clock frequency is higher than the frequency of the three channel signals. The three channels are the first angular displacement sensor channel, the second angular displacement sensor channel, and the linear displacement sensor channel. Based on the first and second count data, an interpolation algorithm is used to determine the equivalent values ​​of each channel signal at the trigger edge signal time, which are then used as the processed table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data.

3. The online detection system for the helix angle error of a gear hobbing machine actuator according to claim 1, characterized in that, The current primary time base is dynamically switched based on the channel stability scores of each channel, including: The system acquires the historical trigger edge signal time interval of the three channels, the historical channel signals of the three channels, the equivalent values ​​of each historical channel signal at the historical main time base trigger edge signal time, and the actual sampled values ​​within the equivalent value calculation window. The historical trigger interval variance of each channel is determined based on the historical trigger edge signal time interval of the three channels; The root mean square of the interpolation alignment residuals of each channel is determined based on the equivalent values ​​of each historical channel signal at the signal time of the historical master time base trigger edge and the actual sampled values ​​within the equivalent value calculation window. Based on the historical channel signals of the three channels, determine the signal-to-noise ratio and the bit loss rate per unit time for each channel; The channel stability score is evaluated based on the trigger interval variance, root mean square of the interpolation alignment residual, noise ratio, and bit loss rate per unit time for each channel. The current primary time base is determined based on the channel stability score.

4. The online detection system for the helix angle error of a gear hobbing machine actuator according to claim 3, characterized in that, Based on channel stability scores, the current primary time base is determined, including: Candidate channels are determined based on channel stability scores; When the candidate channel is inconsistent with the primary time base channel after the last switch, and the time since the last primary time base switch exceeds a preset time length, it is determined whether the difference between the stability score of the candidate channel in the target time period and the stability score of the primary time base channel after the last switch is continuously greater than the first threshold, and / or whether the difference between the stability score of the candidate channel and the stability score of the primary time base channel after the last switch is greater than the second threshold, and the second threshold is less than the first threshold. If the difference between the stability score of the candidate channel during the target time period and the stability score of the primary time base channel after the last switch is continuously greater than the first threshold, and / or the difference between the stability score of the candidate channel and the stability score of the primary time base channel after the last switch is greater than the second threshold, then the time base of the candidate channel is switched to the primary time base.

5. The online detection system for helix angle error of a gear hobbing machine actuator according to claim 1, characterized in that, After the current primary time base is dynamically switched based on the channel stability scores of each channel, it includes: Select the first target sequence before the switching point and the second target sequence after the switching point. The first target sequence and the second target sequence are sequences that overlap in physical time. Calculate the cross-correlation function between the first target sequence and the second target sequence to determine the time offset between the first target sequence and the second target sequence; Align the second target sequence based on the time offset; Based on the fitting algorithm, determine the relationship model between the first target sequence and the adjusted second target sequence; Based on the relational model, the channel signal after the switching point is corrected.

6. The online detection system for the helix angle error of a gear hobbing machine actuator according to claim 5, characterized in that, Also includes: Based on the first weighted sequence, the first target sequence is weighted, and the values ​​of the first weighted sequence are arranged from high to low. Based on the second weighted sequence, the corrected second target sequence is weighted. The values ​​of the second weighted sequence are arranged from low to high. The sum of the weights of the first weighted sequence and the second weighted sequence at the same physical time is 1.

7. The online detection system for helix angle error of a gear hobbing machine actuator according to claim 1, characterized in that, Also includes: Temperature sensor used to collect the ambient temperature of the gear hobbing machine; An accelerometer is used to collect vibration data from a gear hobbing machine. Based on the processed table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data, determine the helix angle error of the helical gear, including: The processed table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data are corrected based on the ambient temperature and vibration data. Based on the current master time base processed and corrected table angular displacement data, tool holder angular displacement data, and tool post vertical linear displacement data, the helix angle error of the helical gear is determined.

8. The online detection system for helix angle error of a gear hobbing machine actuator according to claim 5, characterized in that, Also includes: The memory is used to store table angular displacement data, tool holder angular displacement data, tool post vertical linear displacement data, time offset, relational model, and helical gear helix angle error.

Images