An automobile part automatic production line real-time monitoring method and system

Abstract

The present application relates to the technical field of electric digital data processing, in particular to a kind of automobile parts automatic production line real-time monitoring method and system, comprising: obtaining the current load signal of the main shaft motor of tapping machine on automobile parts production line;And the current load signal is carried out variational modal decomposition and obtains multiple intrinsic modal components, the intrinsic modal component with highest frequency is taken as high-frequency modal component;Current load signal is divided into multiple sampling windows, the fluctuation degree and the jamming degree of high-frequency modal component in each sampling window are calculated;The penalty factor after correction when the next sampling window is carried out variational modal decomposition is calculated, and the jamming degree of the next sampling window is obtained based on the penalty factor after correction;When the jamming degree exceeds the set threshold in continuous multiple sampling windows, it is determined that tapping machine on automobile parts production line exists exception.The present application solves the problem that monitoring accuracy is not high.

Description

Technical Field

[0001] This invention relates to the field of electronic digital data processing technology. More specifically, this invention relates to a real-time monitoring method and system for automated production lines of automotive parts. Background Technology

[0002] As the global automotive industry undergoes a profound transformation towards Industry 4.0 and intelligent manufacturing, the production of automotive parts has long since moved beyond traditional manual operations, adopting fully automated production lines characterized by high precision, high efficiency, and high integration. Within these complex production line systems, precision hole machining (such as tapping, drilling, and reaming) is an indispensable and critical process in the manufacturing of core components like engine blocks, transmission housings, and chassis connectors. Among these, automated tapping machines, as the core equipment for internal thread machining, use a spindle motor to drive a tap to rotate and cut inside the workpiece. This process is characterized by high operating frequency, rapid load changes, large cutting torque, and intermittent operation.

[0003] However, under long-term high-load continuous operation, tapping machines are prone to tap wear, chipping, breakage, and chip blockage due to the combined effects of various complex factors such as uneven workpiece material, cutting fluid lubrication failure, poor chip removal, or tool fatigue. If these faults occur and are not detected in time, they can lead to serious equipment accidents, such as the scrapping of expensive workpiece threads, increasing production costs, or even the breaking of the tap and its jamming inside expensive parts, causing unplanned downtime of the entire automated production line and resulting in huge economic losses and production capacity pressure for automotive parts manufacturers. In the existing monitoring technology system, compared with vibration sensors and acoustic emission sensors, monitoring methods based on spindle motor current signals have become a research hotspot due to their non-invasiveness, low cost, and ease of integration. Among them, variational mode decomposition (VMD), with its solid mathematical theoretical foundation, can effectively overcome the mode aliasing and endpoint effects problems of empirical mode decomposition and is considered a powerful tool for processing non-stationary and nonlinear current load signals.

[0004] However, existing monitoring schemes based on variational mode decomposition generally have the limitation of setting the penalty factor with a fixed empirical value, which makes it difficult to adapt to the complex and ever-changing working conditions of tapping machines. If the fixed penalty factor is set too high, it will forcibly compress the modal bandwidth, thereby incorrectly filtering out the broadband impulse characteristics rich in early faults, resulting in missed alarms. Conversely, if it is set too low, it will introduce a large amount of on-site electromagnetic and mechanical background noise due to the excessively wide bandwidth, resulting in modal aliasing and a sharp drop in the signal-to-noise ratio, causing false alarms and thus leading to low monitoring accuracy. Summary of the Invention

[0005] To address the problem of low monitoring accuracy mentioned in the background art, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a real-time monitoring method for an automated production line for automotive parts, comprising: acquiring current load signals of the spindle motor of a tapping machine on the automotive parts production line at multiple historical time nodes; performing variational mode decomposition on the current load signals to obtain multiple intrinsic mode components, and taking the intrinsic mode component with the highest frequency as the high-frequency mode component; dividing the current load signals into multiple sampling windows, and calculating the fluctuation degree and congestion degree of the high-frequency mode component in each sampling window; the congestion degree is positively correlated with the fluctuation degree and high-frequency mode spread entropy value of the high-frequency mode component in each sampling window, and negatively correlated with the mean of the high-frequency mode spread entropy values ​​in all sampling windows; calculating a corrected penalty factor when performing variational mode decomposition in the next sampling window, the corrected penalty factor being positively correlated with a preset penalty factor of the current sampling window and negatively correlated with the congestion degree of the high-frequency mode component in the current sampling window; obtaining the congestion degree of the next sampling window based on the corrected penalty factor; and determining that there is an anomaly in the tapping machine on the automotive parts production line when the congestion degree exceeds a set threshold in multiple consecutive sampling windows.

[0007] The above technical solution decomposes historical signals to extract high-frequency features and simultaneously calculates a comprehensive blockage index composed of volatility and dispersion within a sliding window framework. Then, it dynamically adjusts the subsequent decomposition process by combining an adaptively updated penalty factor, enabling the algorithm to continuously enhance the ability to identify abnormal features in the time series. This achieves timely and reliable identification of abnormal states of tapping machines, thereby significantly improving the operational safety and intelligent monitoring capabilities of automated production lines for automotive parts.

[0008] Furthermore, the first The degree of fluctuation of high-frequency modal components within each sampling window for: , For the first Within the sampling window, the first High-frequency mode amplitude of each data point For the first The mean of the high-frequency mode amplitudes of all data points within a sampling window For the absolute value function, For the first The total number of data points within each sampling window For the first The standard deviation of the high-frequency mode amplitude of all data points within a sampling window To set the standard deviation of the high-frequency modal amplitude of all data points within the sampling window for the tapping machine in the air-cutting state on the automotive parts production line.

[0009] The aforementioned technical solution compares the discrete fluctuations of high-frequency components within the current window with the stability level under no-load conditions, and comprehensively considers the degree to which all data points within the window deviate from the average amplitude. This allows for the simultaneous representation of both local fluctuation intensity and overall vibration pattern changes. Compared to methods relying solely on a single statistical quantity, this approach more fully reveals the subtle manifestations of abnormal factors such as changes in machining load and tool obstruction in high-frequency signals. Consequently, it significantly improves the sensitivity and reliability of abnormal fluctuation identification, enabling a more accurate characterization and timely warning of the spindle's operating status.

[0010] Furthermore, the first The degree of blockage of high-frequency modal components within each sampling window for: , For the first The degree of fluctuation of high-frequency modal components within each sampling window For normalization function, For the first The spread entropy value of high-frequency modal components within each sampling window To preset the maximum spread entropy value, This represents the average distribution entropy of high-frequency modal components within multiple sampling windows of a tapping machine on an automotive parts production line under normal operating conditions.

[0011] The aforementioned technical solution couples the fluctuation intensity of high-frequency components with the abnormal deviation of dispersion entropy, and through normalization processing, it achieves a unified scale across different windows and operating conditions, thereby forming a comprehensive quantitative indicator that is more sensitive to blockage phenomena. This not only captures hidden anomalies caused by disordered signal energy distribution but also amplifies the differences before and after blockage formation by combining instantaneous fluctuation changes, effectively improving the identifiability of abnormal signs and allowing minute blockage trends in the machining process to be exposed early, thus enhancing the accuracy and early warning capabilities of spindle operating status monitoring.

[0012] Furthermore, the first Penalty factor after variational mode decomposition correction for each sampling window , For the first Preset penalty factor when performing variational mode decomposition within a sampling window For the natural constant An exponential function with base 0. For the first The degree of blockage of high-frequency modal components within each sampling window. , These represent the cumulative mean and cumulative standard deviation of the blockage degree of high-frequency modal components in multiple sampling windows of a tapping machine on an automotive parts production line under normal operating conditions.

[0013] The aforementioned technical solution dynamically adjusts the penalty factor of the previous window exponentially based on the deviation of the current window's congestion level from the normal statistical level. This allows variational mode decomposition to adaptively enhance its ability to separate anomalous components when faced with abnormal signals, while maintaining high steady-state consistency when the signal is in a stable state. This continuous correction mechanism based on historical statistical benchmarks enables the algorithm to respond in real time to changes in congestion levels, improving its sensitivity to anomalous patterns and decomposition accuracy, thereby significantly enhancing the robustness and timeliness of anomaly detection.

[0014] Furthermore, the current load signal specifically refers to: sampling the current of the tapping machine spindle motor during operation using a Hall current sensor or current transformer on the automotive parts production line, converting the sampled current signal into a voltage signal through a signal conditioning circuit, and digitizing the voltage signal to obtain the current load signal.

[0015] Furthermore, the current load signal is divided into multiple sampling windows using a sliding overlapping window method.

[0016] The aforementioned technical solution, by employing a sliding overlapping window, enables continuous signal coverage along the time axis, allowing adjacent windows to share some data. This avoids the loss of abrupt features due to truncation and enhances the ability to capture instantaneous fluctuations and gradual trends. Because the windows overlap, a higher temporal resolution feature sequence can be formed, allowing abnormal changes such as blockages to be amplified and identified at an early stage. This significantly improves the continuity, stability, and sensitivity of the monitoring process, thereby achieving more accurate tracking and timely early warning of the spindle's operating status.

[0017] Furthermore, it also includes using a low-pass filter and an isolation amplifier to perform signal conditioning on the current load signal.

[0018] Furthermore, the preset penalty factor is 2000.

[0019] Furthermore, it also includes standardizing the degree of blockage.

[0020] In a second aspect, the present invention provides a real-time monitoring system for an automated production line of automotive parts, including a memory and a processor. The memory stores computer program instructions, which, when executed by the processor, implement any of the above-described methods for real-time monitoring of an automated production line of automotive parts.

[0021] The beneficial effects of this invention are as follows:

[0022] This invention extracts refined features from the spindle motor current load signal, combining the volatility, dispersion, and deviation of high-frequency components from normal patterns to construct a comprehensive index that sensitively reflects processing anomalies such as blockage and obstruction. Furthermore, it introduces a time-adaptive decomposition penalty factor, enabling the signal decomposition process to automatically enhance the separation capability of abnormal features based on real-time conditions. This allows for the early identification of abnormal trends during tapping machine operation, improving monitoring accuracy, real-time performance, and robustness, thereby significantly enhancing the stable operation and safety assurance capabilities of automated automotive parts production lines. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating a real-time monitoring method for an automated production line of automotive parts according to an embodiment of the present invention;

[0024] Figure 2 This is a schematic diagram illustrating the trend of blockage degree over time according to an embodiment of the present invention;

[0025] Figure 3 This is a schematic block diagram illustrating the structure of a real-time monitoring system for an automated production line of automotive parts according to an embodiment of the present invention. Detailed Implementation

[0026] An embodiment of a real-time monitoring method for an automated production line of automotive parts.

[0027] like Figure 1 As shown in the flowchart, an embodiment of the present invention provides a real-time monitoring method for an automated production line of automotive parts, which includes the following steps:

[0028] S1: Obtain the current load signal of the spindle motor of the tapping machine on the automotive parts production line at multiple historical time points and perform variational mode decomposition to obtain multiple intrinsic mode components. The intrinsic mode component with the highest frequency is taken as the high-frequency mode component.

[0029] In a preferred embodiment, the current load signal is specifically obtained by: sampling the current of the tapping machine spindle motor during operation using a Hall current sensor or current transformer on the automotive parts production line, converting the sampled current signal into a voltage signal through a signal conditioning circuit, and digitizing the voltage signal to obtain the current load signal.

[0030] The current load signal is then subjected to variational mode decomposition to obtain multiple intrinsic mode components, and the intrinsic mode component with the highest frequency is taken as the high-frequency mode component. Variational mode decomposition is a prior art technique and will not be described in detail here.

[0031] Furthermore, the current load signal is conditioned using a low-pass filter and an isolation amplifier. Specifically, the isolation amplifier effectively suppresses common-mode noise and transient high-voltage spike interference commonly found in industrial environments, ensuring the accuracy of monitoring data and the safe operation of the system. The low-pass filter eliminates unwanted high-frequency noise components above the effective bandwidth in the signal and plays a crucial role in anti-aliasing, ensuring the accuracy of subsequent digital processing and the effectiveness of signal analysis. This results in a purer and more reliable reflection of the actual operating conditions of the spindle motor from the extracted load signal.

[0032] S2: Divide the current load signal into multiple sampling windows and calculate the fluctuation and blockage of the high-frequency modal components in each sampling window.

[0033] In a preferred embodiment, the current load signal is divided into multiple sampling windows using a sliding overlapping window method.

[0034] No. The degree of fluctuation of high-frequency modal components within each sampling window for: , For the first Within the sampling window, the first High-frequency mode amplitude of each data point For the first The mean of the high-frequency mode amplitudes of all data points within a sampling window For the absolute value function, For the first The total number of data points within each sampling window For the first The standard deviation of the high-frequency mode amplitude of all data points within a sampling window To set the standard deviation of the high-frequency modal amplitude of all data points within the sampling window for the tapping machine in the air-cutting state on the automotive parts production line.

[0035] By calculating the average absolute deviation of the high-frequency mode amplitudes relative to their mean within the sampling window, the absolute dispersion of the signal is quantified. Simultaneously, using the standard deviation under the empty-cut state as a benchmark, the ratio of the standard deviation of the high-frequency mode amplitudes within the current sampling window to this benchmark is calculated, and this ratio is used as a dynamic weighting coefficient to multiply and weight the aforementioned average absolute deviations. By introducing the empty-cut state as a reference benchmark, the inherent background noise of the equipment is effectively offset. Furthermore, the weighting mechanism using the ratio greatly enhances the sensitivity to non-stationary impact signals. This means that when the tapping machine experiences a slight change in the high-frequency mode amplitude due to a fault, the final calculated fluctuation value exhibits a significant nonlinear amplification effect, thereby greatly improving the signal-to-noise ratio of the fault characteristics and the accuracy of the diagnostic results.

[0036] No. The degree of blockage of high-frequency modal components within each sampling window for: , For the first The degree of fluctuation of high-frequency modal components within each sampling window For normalization function, For the first The spread entropy value of high-frequency modal components within each sampling window To preset the maximum spread entropy value, This represents the average distribution entropy of high-frequency modal components within multiple sampling windows of a tapping machine on an automotive parts production line under normal operating conditions. Furthermore, the degree of blockage is standardized.

[0037] Using the degree of fluctuation as a basis, and introducing the rate of change of the dispersion entropy value after normalization as a dynamic weighting coefficient, the dynamic weighting coefficient comprehensively considers the relative deviation of the current dispersion entropy value from the average dispersion entropy value under normal operating conditions and the proportion relative to the preset maximum dispersion entropy value. This achieves a three-dimensional capture of the characteristics of tapping machine blockage faults. Specifically, it leverages the high sensitivity of dispersion entropy value to signal pattern disorder to enhance the indicative role of fluctuation degree. When blockage occurs, leading to a simultaneous increase in signal amplitude fluctuation and complexity, the calculated index value exhibits a significant nonlinear superposition amplification effect, thereby greatly improving the sensitivity and accuracy of identifying abnormal blockage conditions. Further standardization effectively unifies the data dimensions under different operating conditions, facilitating the subsequent setting of a universal fault threshold.

[0038] like Figure 2 The diagram illustrates the trend of blockage degree over time in an embodiment of the present invention.

[0039] S3: Calculate the corrected penalty factor when performing variational mode decomposition in the next sampling window, and obtain the blocking degree of the next sampling window based on the corrected penalty factor.

[0040] In a preferred embodiment, the first Penalty factor after variational mode decomposition correction for each sampling window , For the first Preset penalty factor when performing variational mode decomposition within a sampling window For the natural constant An exponential function with base 0. For the first The degree of blockage of high-frequency modal components within each sampling window. , These represent the cumulative mean and cumulative standard deviation of the blockage degree of high-frequency modal components of the tapping machine on the automotive parts production line under normal operating conditions within multiple sampling windows. The preset penalty factor is 2000, but it can be set according to actual conditions.

[0041] An inverse adjustment relationship is constructed using an exponential function. This means that the penalty factor for the next sampling window is dynamically adjusted based on the standardized deviation of the congestion level within the current sampling window from the cumulative mean and cumulative standard deviation under normal operating conditions. Specifically, when a significant increase in congestion level is detected, indicating a potential fault, the algorithm automatically reduces the penalty factor, thereby appropriately relaxing the bandwidth constraints of mode decomposition. This allows the algorithm to accommodate and capture the broadband non-stationary impact characteristics caused by the fault, preventing over-smoothing or omission of critical information. Under normal operating conditions, the algorithm maintains a higher penalty factor to ensure excellent frequency separation and noise suppression capabilities, thus achieving adaptive matching of the signal processing process to different operating conditions.

[0042] The degree of congestion in the next sampling window is obtained based on the corrected penalty factor. Specifically, the corrected penalty factor obtained from the current sampling window through the feedback adjustment mechanism is used as a key control variable and applied to the variational mode decomposition operation of the current load signal in the next sampling window. This allows for accurate calculation of the degree of congestion in the next sampling window based on dynamic adjustment of the bandwidth, ensuring that the feature extraction process can match the non-stationary change trend of the signal in real time and avoiding distortion of fault information caused by parameter rigidity.

[0043] S4: In response to the blockage level exceeding a set threshold in multiple consecutive sampling windows, it is determined that there is an abnormality in the tapping machine on the automotive parts production line.

[0044] In a preferred embodiment, by introducing strict temporal consistency verification logic, that is, no longer relying on a sudden change in the index at a single moment, but responding to the stringent condition that the degree of blockage continuously and stably exceeds a set threshold in multiple consecutive sampling windows, it is finally determined that there is an abnormality in the tapping machine on the automotive parts production line. This multi-frame continuous confirmation mechanism significantly improves the anti-interference capability of the system, can effectively filter out random noise interference caused by cutting fluid splashing, power grid transient fluctuations or occasional chip resistance, and ensure that the final alarm signal has extremely high confidence and robustness, thereby accurately locking the real equipment fault state.

[0045] This invention extracts and decomposes high-frequency features from the current load signal of the tapping machine spindle motor, uses a sliding overlapping window strategy to continuously monitor the signal, and constructs a blockage level using volatility and dispersion indicators. An adaptively corrected penalty factor dynamically adjusts the subsequent decomposition process, enhancing the identification of abnormal signal components. Signal conditioning and standardization ensure the stability of the collected data and the consistency of the indicators. Combined with a continuous window threshold determination mechanism, it achieves sensitive detection and timely response to minor blockages and load anomalies. The overall solution effectively distinguishes between normal and abnormal fluctuations, improving the accuracy, robustness, and real-time performance of tapping machine operation status monitoring, and significantly enhancing the safety and intelligent management level of automated automotive parts production lines.

[0046] An example of a real-time monitoring system for an automated production line of automotive parts:

[0047] like Figure 3 As shown in the figure, a structural block diagram of a real-time monitoring system for an automated production line of automotive parts according to an embodiment of the present invention includes a processor and a memory.

[0048] This invention also provides a real-time monitoring system for automated production lines of automotive parts. For example... Figure 3 As shown, the system includes a processor and a memory, the memory storing computer program instructions, which, when executed by the processor, implement the real-time monitoring method for an automated production line of automotive parts according to the present invention.

[0049] The aforementioned real-time monitoring system for an automated production line of automotive parts also includes other components well-known to those skilled in the art, such as communication interfaces. Their settings and functions are known in the art and will not be described in detail here.

[0050] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions stored or otherwise maintained by such a computer-readable medium.

[0051] In the description of this specification, "multiple" or "several" means at least two, such as two, three or more, unless otherwise explicitly specified.

[0052] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A real-time monitoring method for an automated production line of automotive parts, characterized in that, include: The current load signal of the spindle motor of the tapping machine on the automotive parts production line is acquired at multiple historical time points; and the current load signal is subjected to variational mode decomposition to obtain multiple intrinsic mode components, and the intrinsic mode component with the highest frequency is taken as the high-frequency mode component. The current load signal is divided into multiple sampling windows, and the fluctuation and blockage of the high-frequency mode components in each sampling window are calculated. No. The degree of fluctuation of high-frequency modal components within each sampling window for: , For the first Within the sampling window, the first High-frequency mode amplitude of each data point For the first The mean of the high-frequency mode amplitudes of all data points within a sampling window For the absolute value function, For the first The total number of data points within each sampling window For the first The standard deviation of the high-frequency mode amplitude of all data points within a sampling window To set the standard deviation of the high-frequency modal amplitude of all data points within the sampling window for a tapping machine in the air-cutting state on an automotive parts production line; No. The degree of blockage of high-frequency modal components within each sampling window for: , For normalization function, For the first The spread entropy value of high-frequency modal components within each sampling window To preset the maximum spread entropy value, The mean value of the dispersion entropy of high-frequency modal components of a tapping machine on an automotive parts production line under normal operating conditions within multiple sampling windows; Calculate the corrected penalty factor when performing variational mode decomposition in the next sampling window, the th Penalty factor after variational mode decomposition correction for each sampling window , For the first Preset penalty factor when performing variational mode decomposition within a sampling window For the natural constant An exponential function with base 0. , These represent the cumulative mean and cumulative standard deviation of the blockage degree of high-frequency modal components of tapping machines on automotive parts production lines under normal operating conditions within multiple sampling windows. The degree of congestion in the next sampling window is obtained based on the corrected penalty factor; in response to the degree of congestion exceeding a set threshold in multiple consecutive sampling windows, it is determined that there is an abnormality in the tapping machine on the automotive parts production line.

2. The real-time monitoring method for an automated production line of automotive parts according to claim 1, characterized in that, The current load signal is specifically obtained by sampling the current of the tapping machine spindle motor during operation using a Hall current sensor or current transformer on the automotive parts production line, converting the sampled current signal into a voltage signal through a signal conditioning circuit, and then digitizing the voltage signal to obtain the current load signal.

3. The real-time monitoring method for an automated production line of automotive parts according to claim 1, characterized in that, The current load signal is divided into multiple sampling windows using a sliding overlapping window method.

4. The real-time monitoring method for an automated production line of automotive parts according to claim 1, characterized in that, It also includes using a low-pass filter and an isolation amplifier to condition the current load signal.

5. The real-time monitoring method for an automated production line of automotive parts according to claim 1, characterized in that, The preset penalty factor is 2000.

6. The real-time monitoring method for an automated production line of automotive parts according to claim 1, characterized in that, It also includes standardizing the degree of blockage.

7. A real-time monitoring system for an automated production line of automotive parts, characterized in that, The system includes a memory and a processor. The memory stores computer program instructions, which, when executed by the processor, implement the real-time monitoring method for an automated production line of automotive parts as described in any one of claims 1 to 6.

Images