Visual remote abnormity early warning method and system for precision manufacturing process

Abstract

The invention relates to the technical field of equipment fault monitoring, and discloses a visual remote abnormity early warning method and system for a precision manufacturing process, and the method comprises the steps: collecting the quality evidence data and energy consumption evidence data of operation equipment; respectively analyzing functional pleasure and system activation corresponding to the quality evidence data and the energy consumption evidence data; a target quadrant area with the function pleasure corresponding to the system activation degree is inquired in a preset equipment situation ring, and the equipment situation ring comprises a high-risk area of an upper left quadrant, a high-power area of an upper right quadrant, a steady-state area of a lower right quadrant and a recession area of a lower left quadrant; and according to the target quadrant region, carrying out visual remote abnormity early warning on the operation equipment. According to the method, a traditional threshold alarm can be upgraded to an interpretable remote state machine, so that the defects that state semantics are lacked, working conditions cannot be distinguished, and alarm fatigue is caused are overcome.

Description

Technical Field

[0001] This invention relates to a visual remote anomaly early warning method and system for precision manufacturing processes, belonging to the field of equipment fault monitoring technology. Background Technology

[0002] Visualized remote anomaly early warning in precision manufacturing processes is one of the key technologies in the fields of intelligent manufacturing and predictive health management (PHM). Visualization refers to presenting equipment status information to remote users intuitively through a graphical interface. Remote anomaly early warning emphasizes the early identification and alerting of potential equipment performance degradation or failures during the manufacturing process in the case of physical space isolation. Currently, research in this field focuses on threshold early warning methods based on multi-sensor data fusion, that is, by collecting multiple physical signals such as vibration, temperature, and current, and setting alarm thresholds for them to achieve anomaly detection.

[0003] However, these existing technologies have significant limitations in practice. Specifically, their shortcomings mainly lie in the interpretability of state recognition. Because they rely on independent thresholds to judge multiple parameters, the warning information generated by the system often only reflects that a specific parameter has exceeded the standard, and it is difficult to intuitively answer the higher-level operation and maintenance question of the overall operating status of the equipment. This limitation means that although on-site personnel can receive alarm signals, they cannot quickly understand the systemic causes and their severity behind the alarms. For example, they cannot distinguish whether the equipment is in a normal state of high-load and high-efficiency operation or an abnormal state of high energy consumption and low output. Therefore, this lack of state interpretability in existing technologies results in a lack of sufficient state semantic support for remote operation and maintenance decisions, thereby affecting the accuracy of warnings and the efficiency of intervention measures. Summary of the Invention

[0004] This invention provides a visual remote anomaly early warning method and system for precision manufacturing processes. Its main purpose is to upgrade traditional threshold alarms to interpretable remote state machines, thereby solving the defects of missing state semantics, inability to distinguish working conditions, and alarm fatigue.

[0005] To achieve the above objectives, the present invention provides a visual remote anomaly early warning method for precision manufacturing processes, comprising: In the precision manufacturing process, quality evidence data and energy consumption evidence data of the operating equipment are collected. The quality evidence data includes servo tracking error, machining accuracy and surface roughness signals, and the energy consumption evidence data includes spindle load current, feed axis servo current, cooling pump power and total vibration of key points. Analyze the functional satisfaction and system activation levels corresponding to the quality evidence data and the energy consumption evidence data, respectively. In the preset device status ring, query the target quadrant region corresponding to the functional satisfaction and the system activation. The device status ring includes the high-risk region in the upper left quadrant, the high-power region in the upper right quadrant, the steady-state region in the lower right quadrant, and the decay region in the lower left quadrant. Based on the target quadrant region, a visual remote anomaly warning is provided for the operating equipment.

[0006] Optionally, the step of analyzing the functional satisfaction and system activation corresponding to the quality evidence data and the energy consumption evidence data respectively includes: The quality evidence data is classified according to rules to obtain the functional satisfaction level; The energy consumption evidence data is classified according to rules to obtain the system activation degree.

[0007] Optionally, the step of classifying the quality evidence data according to rules to obtain functional satisfaction includes: The servo tracking error, machining accuracy, and surface roughness signals in the quality evidence data are extracted by synchronous time windows to obtain the error sequence, dimensional deviation sequence, and noise feature sequence within the window; Calculate the maximum absolute value and standard deviation of the in-window error sequence; Calculate the center drift value of the size deviation sequence; Extract the high-frequency energy percentage of the noise feature sequence; The maximum absolute value, the standard deviation, the center drift value, and the high-frequency energy ratio are input into a preset quality rating table to query the functional satisfaction level in the quality rating table. The quality rating table refers to a two-dimensional lookup matrix with the maximum absolute value level as the row index, the standard deviation level as the column index, and the center drift coefficient and energy proportion coefficient as the downgrade criteria.

[0008] Optionally, the step of classifying the energy consumption evidence data according to rules to obtain the system activation degree includes: The spindle load current, feed axis servo current, cooling pump power, and total vibration at key points in the energy consumption evidence data are extracted by synchronous time windows to obtain current sequence, power sequence, and vibration sequence. Calculate the effective value of the current and the variance of the fluctuation in the current sequence; Calculate the mean and peak-to-peak values ​​of the power sequence; Calculate the total effective value of the vibration sequence; The effective value of the current, the fluctuation variance, the mean, the peak-to-peak value, and the total effective value are input into a preset energy consumption rating table to query the system activation level in the energy consumption rating table; The energy consumption rating table refers to a two-dimensional lookup matrix with effective value level as row index, fluctuation variance level as column index, and cooling power consumption coefficient and vibration total coefficient as upgrade standards.

[0009] Optionally, before querying the target quadrant region corresponding to the functional satisfaction and system activation in the preset device status loop, the method further includes: A two-dimensional rectangular coordinate system is established with the functional pleasure level of historical periods as the horizontal axis and the system activation level of historical periods as the vertical axis. The positive and negative boundary value of the horizontal axis in the two-dimensional rectangular coordinate system is set as the functional pleasure threshold, and the positive and negative boundary value of the vertical axis in the two-dimensional rectangular coordinate system is set as the system activation threshold. Based on the functional pleasure threshold and the system activation threshold, the first quadrant, the second quadrant, the third quadrant and the fourth quadrant in the two-dimensional rectangular coordinate system are respectively regarded as the high-performance region of the upper right quadrant, the high-risk region of the upper left quadrant, the decay region of the lower left quadrant and the steady-state region of the lower right quadrant. The target quadrant region is determined by the two-dimensional rectangular coordinate system, the high-power region in the upper right quadrant, the high-risk region in the upper left quadrant, the decay region in the lower left quadrant, and the steady-state region in the lower right quadrant.

[0010] Optionally, the step of providing visual remote anomaly warning for the operating equipment based on the target quadrant region includes: Render the device status loop on the remote monitoring interface to obtain the rendered status loop; The dynamic emotion points corresponding to functional pleasure and system activation in the target quadrant are displayed in real time within the rendering situation ring. By measuring the duration of the dynamic emotion point in the target quadrant, a high-risk anomaly warning for the operating equipment is generated, thereby enabling a visual remote anomaly warning for the operating equipment. By analyzing the movement trajectory of the dynamic emotion points between different target quadrants, a trend anomaly warning for the operating equipment is generated, thereby enabling visualized remote anomaly warning for the operating equipment.

[0011] Optionally, generating an anomaly warning for the operating device based on the movement trajectory of the dynamic emotion point between different target quadrants includes: When the dynamic emotion point switches from the steady-state area in the lower right quadrant to the decay area in the lower left quadrant, a performance degradation warning for the running device is generated. When the dynamic emotion point switches from the steady-state area in the lower right quadrant and the decay area in the lower left quadrant to the high-risk area in the upper left quadrant, an abnormal change warning is generated for the operating equipment. The performance degradation warning and the abnormal mutation warning are used to determine the trend anomaly warning of the operating equipment.

[0012] Optionally, after performing visual remote anomaly warning for the operating equipment based on the target quadrant region, the method further includes: Cluster analysis was performed on dynamic emotion points over a continuous period of time to obtain an emotion scatter plot; Calculate the first distribution area of ​​the steady-state region in the lower right quadrant of the emotion scatter plot; Calculate the second distribution area of ​​the high-risk area in the upper left quadrant and the decay area in the lower left quadrant of the emotion scatter plot; The ratio of the first distribution area to the second distribution area is used as the health decline index; Preventive maintenance recommendations for the operating equipment are generated using the health decline index.

[0013] Optionally, the collection of quality evidence data and energy consumption evidence data of the operating equipment includes: During each control cycle, the servo tracking error of the operating equipment is collected through the CNC servo interface; The machining accuracy of the operating equipment is obtained through the in-machine probe of the operating equipment. The surface roughness signal of the operating device is extracted through the microphone at the front end of the spindle of the operating device. The servo tracking error, the machining accuracy, and the surface roughness signal are used as quality evidence data. During each control cycle, the spindle load current and feed axis servo current of the operating device are collected through the driver current loop register; The cooling pump power of the operating equipment is collected using a bus power meter; The total vibration of key points of the operating equipment is collected using an accelerometer. The spindle load current, the feed axis servo current, the cooling pump power, and the total vibration at the key points are used as energy consumption evidence data.

[0014] To address the aforementioned problems, the present invention also provides a visual remote anomaly early warning system for precision manufacturing processes, the system comprising: The data collection module is used to collect quality evidence data and energy consumption evidence data of the operating equipment during the precision manufacturing process. The quality evidence data includes servo tracking error, machining accuracy and surface roughness signals, and the energy consumption evidence data includes spindle load current, feed axis servo current, cooling pump power and total vibration of key points. The coordinate calculation module is used to analyze the functional satisfaction and system activation corresponding to the quality evidence data and the energy consumption evidence data, respectively. The quadrant query module is used to query the target quadrant region corresponding to the functional satisfaction and the system activation in a preset device status ring. The device status ring includes the high-risk region in the upper left quadrant, the high-power region in the upper right quadrant, the steady-state region in the lower right quadrant, and the decay region in the lower left quadrant. The anomaly warning module is used to provide visual remote anomaly warnings for the operating equipment based on the target quadrant area.

[0015] Compared to the problems described in the background art, the embodiments of the present invention, by collecting quality evidence data and energy consumption evidence data of operating equipment, construct a binary perspective describing the equipment status from the data source. This overcomes the problem of isolated multiple parameters in the background art, laying a unified and complementary data foundation for subsequent comprehensive evaluation of whether the equipment is efficiently producing or inefficiently idling. The collected data is directly related to processing results and energy consumption, enabling subsequent analysis to directly reveal the intrinsic relationship between quality and energy consumption. Furthermore, the embodiments of the present invention integrate and elevate a series of isolated, low-level physical parameters into two high-level indicators with clear business meanings: using the functional satisfaction indicator to characterize whether the work is done well and using the system activation indicator to characterize whether the work is laborious. This directly solves the problem in the background art of not being able to intuitively answer the overall equipment status issue. The invention addresses the shortcomings of traditional threshold alarms by providing operators with easily understandable state descriptions, significantly enhancing interpretability. Furthermore, by mapping two-dimensional indicators to four semantically meaningful quadrants—high-risk zone, high-power zone, steady-state zone, and decay zone—the complex health status of equipment is transformed into a clear, categorized situational diagnosis. This not only resolves the confusion in the prior art regarding the inability to distinguish between high-load, high-efficiency operation and high-energy-consumption, low-output, but also allows remote users to instantly identify the most likely abnormal mode or health state of the current equipment, providing a clear basis for decision-making. Moreover, the invention ultimately achieves state-based, interpretable early warnings. Unlike traditional low-level alarms such as threshold exceeding limits, the system can now issue higher-level warnings such as "Equipment is in a high-risk zone, please pay attention!" or "Equipment shows a decay trend." This warning information is rich in semantics, directly revealing the nature of the anomaly, thus strongly supporting remote operation and maintenance decisions, making intervention measures more targeted, and ultimately improving the accuracy of warnings and the efficiency of operation and maintenance. Therefore, this invention can upgrade traditional threshold alarms to interpretable remote state machines, thereby solving the defects of missing state semantics, inability to distinguish operating conditions, and alarm fatigue. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a visual remote anomaly early warning method for precision manufacturing processes provided in an embodiment of the present invention. Figure 2A rendered situation diagram of a visualization-based remote anomaly early warning method for precision manufacturing processes provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of a module for implementing the visual remote anomaly early warning system for precision manufacturing processes, provided as an embodiment of the present invention.

[0017] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] This application provides a method for visual remote anomaly early warning in precision manufacturing processes. The executing entity of this method includes, but is not limited to, at least one of the following computer devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0020] Reference Figure 1 The diagram shown is a flowchart illustrating a visual remote anomaly early warning method for precision manufacturing processes according to an embodiment of the present invention. In this embodiment, the visual remote anomaly early warning method for precision manufacturing processes includes: S1. During the precision manufacturing process, collect quality evidence data and energy consumption evidence data of the operating equipment. The quality evidence data includes servo tracking error, machining accuracy and surface roughness signals. The energy consumption evidence data includes spindle load current, feed axis servo current, cooling pump power and total vibration of key points.

[0021] This invention collects quality evidence data and energy consumption evidence data of operating equipment to construct a binary perspective describing the equipment status from the data source. This overcomes the problem of isolated multiple parameters in the prior art and lays a unified and complementary data foundation for subsequent comprehensive evaluation of whether the equipment is producing efficiently or idling inefficiently. The collected data is directly related to the processing results and energy consumption, enabling subsequent analysis to directly reveal the intrinsic relationship between quality and energy consumption.

[0022] The precision manufacturing process refers to machining processes requiring micron-level precision, such as CNC riveting, milling and turning, precision grinding, micro-stamping, laser micromachining, and micro-electrical discharge machining. The operating equipment refers to a CNC machine tool platform integrating a spindle, tool holder, cutting tool, and coolant system, such as a lathe, milling machine, grinding machine, machining center, five-axis CNC machine tool, and precision riveting machine. The quality evidence data refers to real-time process information used to characterize the workpiece size and surface quality, mainly including servo tracking error, machining accuracy, and surface roughness signals. The energy consumption evidence data refers to real-time status information used to characterize the equipment's energy consumption and mechanical stress, mainly including spindle load current, feed axis servo current, coolant pump power, and total vibration at key points.

[0023] In one embodiment of the present invention, the collection of quality evidence data and energy consumption evidence data of the operating equipment includes: acquiring the servo tracking error of the operating equipment through a CNC servo interface in each control cycle; obtaining the machining accuracy of the operating equipment through an in-machine probe of the operating equipment; extracting the surface roughness signal of the operating equipment through a microphone at the front end of the spindle of the operating equipment; using the servo tracking error, the machining accuracy, and the surface roughness signal as quality evidence data; acquiring the spindle load current and feed axis servo current of the operating equipment through a driver current loop register in each control cycle; acquiring the cooling pump power of the operating equipment through a bus power meter; acquiring the total vibration of key points of the operating equipment through an accelerometer; and using the spindle load current, the feed axis servo current, the cooling pump power, and the total vibration of key points as energy consumption evidence data.

[0024] The control cycle refers to the fixed clock tick of the servo loop of the control system, such as a 1ms servo interrupt cycle, a 0.5ms position loop cycle, and a 2ms speed loop cycle. The CNC servo interface refers to the open real-time servo data communication port of the CNC system, such as the position error register, speed error register, current feedback register, D / A output interface, and PDO fieldbus port. The servo tracking error refers to the instantaneous difference between the position command value calculated by the CNC system in each servo cycle and the actual feedback value, such as the ADC code value read from the position error register, the analog voltage output from the speed error register, and the contour error data in the PDO fieldbus packet. The machining accuracy refers to the original measurement obtained after the in-machine probe repeatedly touches the critical dimensions of the workpiece. The data sequence refers to raw sequences of data, such as multiple aperture touch values, step height scan values, and flatness point cloud data, without statistical processing. The surface roughness signal refers to the raw cutting audio stream collected by the microphone at the front end of the spindle without feature extraction, such as PCM encoded audio streams, pulse code modulation signals, and raw sampling point sequences. The driver current loop register refers to the hardware register unit inside the servo driver used to store instantaneous current feedback code values, such as raw digital quantities storing phase current ADC code values, q-axis current count values, d-axis current count values, and current percentage code values. The spindle load current refers to the current recorded in the spindle servo driver current loop register. The raw digital quantities characterizing motor torque include, for example, instantaneous phase current ADC code values, q-axis current counting sequences, and other raw data. The feed axis servo current refers to the raw digital quantities characterizing the feed cutting force recorded in the current loop register of the X / Y / Z feed axis servo driver, such as instantaneous phase current ADC code values, q-axis current counting sequences, and raw sampling point sequences. The bus power meter refers to an intelligent electrical measurement module connected to the fieldbus, such as a Modbus-RTU power meter, CAN-open power module, Profinet power sensor, EtherCAT power module, and other hardware. The coolant pump power refers to the aforementioned bus power... The unaveraged voltage-current product raw count inside the accelerometer, such as the product count of instantaneous voltage and current sample values, the raw accumulated code value of the power calculation unit, etc. The accelerometer refers to a vibration sensor mounted on the spindle bearing housing or feed slide, such as a MEMS accelerometer, piezoelectric accelerometer, capacitive accelerometer, fiber optic accelerometer, or other sensor device or probe body. The total vibration at the key point refers to the raw sampling sequence output from the above-mentioned accelerometer port without integration, filtering, or RMS value calculation, such as the raw digital code value sequence of the X / Y / Z axis, the analog voltage sampling sequence after primary amplification, the raw pulse count sequence output from the charge amplifier, etc.

[0025] S2. Analyze the functional satisfaction and system activation corresponding to the quality evidence data and the energy consumption evidence data respectively.

[0026] This invention integrates and elevates a series of isolated, low-level physical parameters into two high-level indicators with clear business meanings: functional satisfaction to characterize whether the work is done well and system activation to characterize whether the work is laborious. This directly solves the problem in the background technology that it is impossible to intuitively answer what state the overall equipment is in, provides operators with an easy-to-understand state description, and greatly enhances the interpretability of the state.

[0027] In one embodiment of the present invention, the step of analyzing the functional satisfaction and system activation corresponding to the quality evidence data and the energy consumption evidence data respectively includes: classifying the quality evidence data according to rules to obtain functional satisfaction; and classifying the energy consumption evidence data according to rules to obtain system activation.

[0028] In another embodiment of the present invention, the step of classifying the quality evidence data according to rules to obtain functional satisfaction includes: extracting the servo tracking error, machining accuracy, and surface roughness signals in the quality evidence data into synchronous time windows to obtain an in-window error sequence, a size deviation sequence, and a noise feature sequence; calculating the maximum absolute value and standard deviation of the in-window error sequence; calculating the center drift value of the size deviation sequence; extracting the high-frequency energy proportion of the noise feature sequence; and inputting the maximum absolute value, the standard deviation, the center drift value, and the high-frequency energy proportion into a preset quality scoring table to query the functional satisfaction in the quality scoring table; wherein, the quality scoring table is a two-dimensional lookup table matrix with the maximum absolute value level as the row index, the standard deviation level as the column index, and the center drift coefficient and the energy proportion coefficient as the degradation criteria.

[0029] The term "intra-window error sequence" refers to the raw code value sequence obtained by truncating the servo tracking error within a fixed time window, without any statistical calculations. The term "dimensional deviation sequence" refers to the unstatistically measured measurement value sequence obtained by truncating the raw machining accuracy measurement data within the same time window. The term "noise feature sequence" refers to the sampling point sequence obtained by truncating the raw surface roughness audio stream within the same time window without feature extraction. The term "maximum absolute value" refers to the instantaneous maximum amplitude obtained from the first statistical analysis of the intra-window error sequence. The term "standard deviation" refers to the first-order dispersion obtained from the first statistical analysis of the error sequence within the same window. The term "center drift value" refers to the absolute deviation between the mean obtained from the first statistical analysis of the dimensional deviation sequence and the tolerance band center. The term "high-frequency energy ratio" refers to the 0-1 relative quantity obtained from the first feature extraction of the noise feature sequence. The term "quality scoring table" refers to the two-dimensional lookup table matrix used for the first time in this step, with row indices representing the maximum absolute value level, column indices representing the standard deviation level, and cells containing the basic pleasure score. After downgrading and correcting the center drift coefficient and energy proportion coefficient, the final functional satisfaction level is output. It should be noted that the construction method of the quality scoring table includes, but is not limited to, the following: Taking the in-window error sequence, size deviation sequence, and noise characteristic sequence from historical normal operation periods, calculating the maximum absolute value, standard deviation, center drift value, and high-frequency energy proportion within the same window to form a quadruplet sample library; dividing the maximum absolute value and standard deviation into equal intervals or equal sample levels, classifying them into 0-9 levels, and obtaining a two-dimensional discrete level boundary. The center drift value and energy ratio are divided into two levels, normal and excessive, using a single threshold. A two-dimensional matrix is ​​established with the maximum absolute value level as the row and the standard deviation level as the column. The top left corner (0,0) is filled with 100 points, which means that the error is small and stable, which is the most pleasant. From the top left to the bottom right main diagonal, i.e. (0,0) to (9,9), the score is reduced by 10 for each square that moves to the bottom right, until the bottom right corner (9,9) is filled with 0 points, which means that the error is large and the fluctuation is large, which is the worst. If the center drift exceeds the standard or the energy ratio exceeds the standard, the base score is reduced by 10 points.

[0030] Furthermore, to better understand the specific content of the quality scoring table in the above-described visual remote anomaly early warning method for precision manufacturing processes, please refer to Table 1 below, which is a table showing the specific content of the quality scoring table in the visual remote anomaly early warning method for precision manufacturing processes provided by an embodiment of the present invention.

[0031] Table 1 Quality Scoring Table

[0032] Continued from Table 1 Quality Scoring Table

[0033] Table 1 above clearly shows the distribution of pleasure levels for different functions in the quality rating table.

[0034] Optionally, the process of inputting the maximum absolute value, the standard deviation, the center drift value, and the high-frequency energy ratio into a preset quality rating table to query the functional pleasure level in the quality rating table is as follows: the maximum absolute value and standard deviation obtained in the current window are mapped to integer levels according to preset grading boundaries, such as levels 0-9. The mapping method is as follows: the minimum and maximum absolute values ​​of the historical period are divided into 10 intervals, which correspond to levels 0-9. The current maximum absolute value is queried to find which interval it falls into, and the corresponding level can be obtained. The standard deviation and the maximum absolute value are similarly processed to form row index and column index. The two-dimensional rating matrix is ​​queried using this (row, column) combination to read the corresponding basic pleasure score (0-100). If the center drift value is greater than the drift threshold or the high-frequency energy ratio is greater than the energy threshold, the basic score is reduced by one level, that is, reduced by 10 points. The corrected score is then graded again to output the final functional pleasure level.

[0035] In another embodiment of the present invention, the step of classifying the energy consumption evidence data according to rules to obtain the system activation degree includes: extracting the spindle load current, feed axis servo current, cooling pump power, and total vibration of key points from the energy consumption evidence data into a synchronous time window to obtain a current sequence, a power sequence, and a vibration sequence; calculating the effective value and fluctuation variance of the current sequence; calculating the mean and peak-to-peak value of the power sequence; calculating the total effective value of the vibration sequence; and inputting the effective value of the current, the fluctuation variance, the mean, the peak-to-peak value, and the total effective value into a preset energy consumption scoring table to query the system activation degree in the energy consumption scoring table; wherein, the energy consumption scoring table is a two-dimensional lookup table matrix with the effective value level as the row index, the fluctuation variance level as the column index, and the cooling power consumption coefficient and the total vibration coefficient as the upgrade standard.

[0036] The current sequence refers to the instantaneous sampling sequence obtained by truncating the raw code values ​​of the spindle load current and feed axis servo current within the same time window without calculating the effective value; the power sequence refers to the instantaneous sampling sequence obtained by truncating the raw product count of the coolant pump power within the same time window without averaging; the vibration sequence refers to the instantaneous sampling sequence obtained by truncating the raw sampling points of the total vibration of key points within the same time window without calculating the effective value; the effective value of the current refers to the root mean square value obtained by performing the first statistical analysis on the current sequence; and the fluctuation variance refers to the variance obtained by performing the first statistical analysis on the current sequence within the same window. The statistically obtained first-order dispersion, the mean refers to the arithmetic mean obtained from the first statistical analysis of the power sequence, the peak-to-peak value refers to the difference between the maximum and minimum values ​​obtained from the first statistical analysis of the power sequence within the same window, the total effective value refers to the root mean square acceleration obtained from the first statistical analysis of the vibration sequence, the energy consumption score table refers to the two-dimensional lookup table matrix used for the first time in the current step, the row index is the current effective value level, the column index is the fluctuation variance level, the cell contains the basic activation score, and after upgrading and correction by the mean, the peak-to-peak value and the total effective value, the final system activation level is output.

[0037] Optionally, the process of inputting the current effective value, the fluctuation variance, the mean, the peak-to-peak value, and the total effective value into a preset energy consumption scoring table to query the system activation level in the energy consumption scoring table is as follows: using the current effective value as the row index and the fluctuation variance as the column index, a two-dimensional scoring matrix is ​​queried using (row, column) combination to read the corresponding basic activation level score (0-100). If the mean is greater than the mean threshold, the peak-to-peak value is greater than the peak-to-peak threshold, or the total effective value is greater than the vibration threshold, the basic activation level score is increased by one level, that is, increased by 10 points. The corrected score is then divided into levels again, and the final activation level is output.

[0038] S3. In the preset device status ring, query the target quadrant region corresponding to the functional satisfaction and the system activation. The device status ring includes the high-risk region in the upper left quadrant, the high-power region in the upper right quadrant, the steady-state region in the lower right quadrant, and the decay region in the lower left quadrant.

[0039] This invention maps two-dimensional indicators to four intuitively semantic quadrants: high-risk zone, high-power zone, steady-state zone, and decay zone. This transforms complex equipment health conditions into a clear and categorized situational diagnosis result. This not only solves the confusion in the prior art regarding the inability to distinguish between high-load, high-efficiency operation and high-energy-consumption, low-output operation, but also allows remote users to quickly identify which abnormal mode or health state the current equipment is most likely to belong to, providing a clear basis for decision-making.

[0040] In one embodiment of the present invention, before querying the target quadrant region corresponding to the functional satisfaction and system activation in a preset device status loop, the method further includes: establishing a two-dimensional rectangular coordinate system with the functional satisfaction of historical time periods as the abscissa and the system activation of historical time periods as the ordinate; setting the positive and negative boundary values ​​of the abscissa in the two-dimensional rectangular coordinate system as the functional satisfaction threshold and the positive and negative boundary values ​​of the ordinate in the two-dimensional rectangular coordinate system as the system activation threshold; based on the functional satisfaction threshold and the system activation threshold, respectively designating the first quadrant, the second quadrant, the third quadrant, and the fourth quadrant in the two-dimensional rectangular coordinate system as the high-power region of the upper right quadrant, the high-risk region of the upper left quadrant, the decay region of the lower left quadrant, and the steady-state region of the lower right quadrant; and determining the target quadrant region through the two-dimensional rectangular coordinate system, the high-power region of the upper right quadrant, the high-risk region of the upper left quadrant, the decay region of the lower left quadrant, and the steady-state region of the lower right quadrant.

[0041] The functional satisfaction threshold refers to the dividing score used for positive and negative division of the horizontal axis after classifying the functional satisfaction scores of historical periods. For example, a score greater than or equal to 60 indicates high satisfaction (positive half-axis), and a score less than 60 indicates low satisfaction (negative half-axis). This score is the functional satisfaction threshold. The system activation threshold refers to the dividing score used for positive and negative division of the vertical axis after classifying the system activation scores of historical periods. For example, a score greater than or equal to 60 indicates high activation (positive half-axis), and a score less than 60 indicates low activation (negative half-axis). This score is the system activation threshold. It should be noted that 60 is just an example; the specific dividing value needs to be determined by taking the median level after classifying the samples from the first N control cycles of normal operation. The high-performance zone in the upper right quadrant refers to the horizontal axis in a two-dimensional rectangular coordinate system. The region in the upper left quadrant, where the x-coordinate is greater than or equal to the functional satisfaction threshold and the y-coordinate is greater than or equal to the system activation threshold, represents a legal heavy-load operation state with high process quality and high energy consumption. The high-risk region in the upper left quadrant, where the x-coordinate is less than the functional satisfaction threshold and the y-coordinate is greater than or equal to the system activation threshold in a two-dimensional rectangular coordinate system, represents an abnormally dangerous state with low process quality and high energy consumption. The decay region in the lower left quadrant, where the x-coordinate is less than the functional satisfaction threshold and the y-coordinate is less than the system activation threshold in a two-dimensional rectangular coordinate system, represents a performance degradation state with low process quality and low energy consumption. The steady-state region in the lower right quadrant, where the x-coordinate is greater than or equal to the functional satisfaction threshold and the y-coordinate is less than the system activation threshold in a two-dimensional rectangular coordinate system, represents a stable light-load operation state with high process quality and low energy consumption.

[0042] S4. Based on the target quadrant region, perform visual remote anomaly early warning for the operating equipment.

[0043] This invention ultimately achieves state-based, interpretable early warnings. Unlike traditional low-level alarms such as threshold exceeding limits, the system can now issue high-level warnings such as "Equipment is in a high-risk area, please pay attention!" or "Equipment shows a deterioration trend." This warning information is rich in semantics, directly revealing the nature of the anomaly, thereby strongly supporting remote operation and maintenance decisions, making intervention measures more targeted, and ultimately improving the accuracy of early warnings and the efficiency of operation and maintenance.

[0044] In one embodiment of the present invention, the step of providing a visual remote anomaly warning for the operating device based on the target quadrant includes: rendering a device status ring on a remote monitoring interface to obtain a rendered status ring; displaying dynamic emotion points corresponding to functional pleasure and system activation in the target quadrant in real time within the rendered status ring; generating a high-risk anomaly warning for the operating device based on the dwell time of the dynamic emotion points in the target quadrant, thereby achieving a visual remote anomaly warning for the operating device; and generating a trend anomaly warning for the operating device based on the movement trajectory of the dynamic emotion points between different target quadrants, thereby achieving a visual remote anomaly warning for the operating device.

[0045] The rendering status ring refers to a colored ring drawn on the remote monitoring interface, with each of the four quadrants colored to carry the real-time display of dynamic emotion points. The dynamic emotion points refer to the coordinate points (functional pleasure, system activation) drawn in real time as points within the rendering status ring, which are refreshed once per control cycle. The dwell time refers to the cumulative time that a dynamic emotion point is continuously in the same quadrant, counted in control cycles, and reset upon entering another quadrant. The movement trajectory refers to the quadrant migration path of the dynamic emotion point between adjacent control cycles, recorded using a quadrant number sequence.

[0046] See Figure 2 The image shown is a rendered situation diagram of a visualized remote anomaly early warning method for precision manufacturing processes provided in an embodiment of the present invention. Figure 2 In the diagram, small white circles represent dynamic emotional points, the horizontal axis represents functional pleasure, and the vertical axis represents system activation. The diagram is divided into the high-risk area in the upper left quadrant, the high-performance area in the upper right quadrant, the steady-state area in the lower right quadrant, and the decline area in the lower left quadrant by the horizontal and vertical axes.

[0047] In one embodiment of the present invention, generating a trend anomaly warning for the operating device based on the movement trajectory of the dynamic emotion point between different target quadrants includes: generating a performance degradation warning for the operating device when the dynamic emotion point switches from the steady-state area of ​​the lower right quadrant to the decay area of ​​the lower left quadrant; generating an abnormal change warning for the operating device when the dynamic emotion point switches from the steady-state area of ​​the lower right quadrant and the decay area of ​​the lower left quadrant to the high-risk area of ​​the upper left quadrant; and determining a trend anomaly warning for the operating device through the performance degradation warning and the abnormal change warning.

[0048] The performance degradation warning is a one-time alert triggered when the trajectory first transitions from the steady-state region to the degradation region, indicating that the device has entered the slow degradation stage. The abnormal change warning is a one-time alert triggered when the trajectory first transitions from the steady-state region to the high-risk region or from the degradation region to the high-risk region, indicating that the device has entered the acute abnormal stage.

[0049] In one embodiment of the present invention, after performing a visual remote anomaly warning for the operating equipment based on the target quadrant region, the method further includes: performing cluster analysis on dynamic emotion points within a continuous time period to obtain an emotion scatter plot; calculating the first distribution area of ​​the steady-state region of the emotion scatter plot in the lower right quadrant; calculating the second distribution area of ​​the high-risk region of the emotion scatter plot in the upper left quadrant and the decline region in the lower left quadrant; using the ratio of the first distribution area to the second distribution area as a health decline index; and generating preventive maintenance recommendations for the operating equipment based on the health decline index.

[0050] The emotion scatter plot refers to a two-dimensional density point set obtained by clustering all dynamic emotion points within a continuous time period, used for area calculation. The first distribution area refers to the point cloud projection area of ​​the emotion scatter plot in the steady-state region of the lower right quadrant, measured in pixels or normalized units. The second distribution area refers to the sum of the point cloud projection areas of the emotion scatter plot in the high-risk region of the upper left quadrant and the decay region of the lower left quadrant. The health decay index is the ratio of the first distribution area to the second distribution area; the smaller the value, the worse the device's health status. The preventive maintenance recommendation refers to the suggestion that when the health decay index falls below a set threshold, this set threshold has been in effect for a historical period. The statistics obtained from the segment, for example, during the period when the equipment has been put into production but has not yet reached its maintenance cycle, are used to calculate the health decline index of machines that require preventive maintenance and the health decline index of machines that do not require preventive maintenance. The boundary value between the two is used as a set threshold. The inspection or maintenance prompts generated by the system only contain text suggestions and do not trigger automatic shutdown, such as "Please arrange preventive maintenance for the equipment". This sentence "Please arrange preventive maintenance for the equipment" is used in conjunction with the abnormal detection in the actual process. For example, if a system is configured in the factory to detect where the machine is malfunctioning, the sentence "Please arrange preventive maintenance for the equipment" can be used to remind the staff to perform maintenance.

[0051] Compared to the problems described in the background art, the embodiments of the present invention, by collecting quality evidence data and energy consumption evidence data of operating equipment, construct a binary perspective describing the equipment status from the data source. This overcomes the problem of isolated multiple parameters in the background art, laying a unified and complementary data foundation for subsequent comprehensive evaluation of whether the equipment is efficiently producing or inefficiently idling. The collected data is directly related to processing results and energy consumption, enabling subsequent analysis to directly reveal the intrinsic relationship between quality and energy consumption. Furthermore, the embodiments of the present invention integrate and elevate a series of isolated, low-level physical parameters into two high-level indicators with clear business meanings: using the functional satisfaction indicator to characterize whether the work is done well and using the system activation indicator to characterize whether the work is laborious. This directly solves the problem in the background art of not being able to intuitively answer the overall equipment status issue. The invention addresses the shortcomings of traditional threshold alarms by providing operators with easily understandable state descriptions, significantly enhancing interpretability. Furthermore, by mapping two-dimensional indicators to four semantically meaningful quadrants—high-risk zone, high-power zone, steady-state zone, and decay zone—the complex health status of equipment is transformed into a clear, categorized situational diagnosis. This not only resolves the confusion in the prior art regarding the inability to distinguish between high-load, high-efficiency operation and high-energy-consumption, low-output, but also allows remote users to instantly identify the most likely abnormal mode or health state of the current equipment, providing a clear basis for decision-making. Moreover, the invention ultimately achieves state-based, interpretable early warnings. Unlike traditional low-level alarms such as threshold exceeding limits, the system can now issue higher-level warnings such as "Equipment is in a high-risk zone, please pay attention!" or "Equipment shows a decay trend." This warning information is rich in semantics, directly revealing the nature of the anomaly, thus strongly supporting remote operation and maintenance decisions, making intervention measures more targeted, and ultimately improving the accuracy of warnings and the efficiency of operation and maintenance. Therefore, this invention can upgrade traditional threshold alarms to interpretable remote state machines, thereby solving the defects of missing state semantics, inability to distinguish operating conditions, and alarm fatigue.

[0052] like Figure 3 The diagram shown is a functional block diagram of a visual remote anomaly early warning system for precision manufacturing processes according to the present invention.

[0053] The visual remote anomaly early warning system 300 for precision manufacturing processes described in this invention can be installed in a computer device. Depending on the functions implemented, the visual remote anomaly early warning system for precision manufacturing processes may include a data collection module 301, a coordinate calculation module 302, a quadrant query module 303, and an anomaly early warning module 304. The module described in this invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by the processor of a computer device and can perform a fixed function, and are stored in the memory of the computer device.

[0054] In this embodiment of the invention, the functions of each module / unit are as follows: The data collection module 301 is used to collect quality evidence data and energy consumption evidence data of the operating equipment during the precision manufacturing process. The quality evidence data includes servo tracking error, machining accuracy and surface roughness signals, and the energy consumption evidence data includes spindle load current, feed axis servo current, cooling pump power and total vibration of key points. The coordinate calculation module 302 is used to analyze the functional satisfaction and system activation corresponding to the quality evidence data and the energy consumption evidence data, respectively. The quadrant query module 303 is used to query the target quadrant region corresponding to the functional satisfaction and the system activation in a preset device status ring. The device status ring includes the high-risk region in the upper left quadrant, the high-power region in the upper right quadrant, the steady-state region in the lower right quadrant, and the decay region in the lower left quadrant. The anomaly warning module 304 is used to provide visual remote anomaly warnings for the operating equipment based on the target quadrant region.

[0055] In detail, the modules in the visual remote anomaly early warning system 300 for precision manufacturing processes described in this embodiment of the invention employ the same methods as described above during use. Figure 1 The method uses the same technical means as the visual remote anomaly early warning method for precision manufacturing processes described in the article, and can produce the same technical effect, so it will not be repeated here.

[0056] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0057] Finally, it should be noted that in the above embodiments, each embodiment can be combined with each other or independent. Deleting any one of them will not affect the technical implementation of other embodiments. The above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for visual remote anomaly early warning in precision manufacturing processes, characterized in that, The method includes: In the precision manufacturing process, quality evidence data and energy consumption evidence data of the operating equipment are collected. The quality evidence data includes servo tracking error, machining accuracy and surface roughness signals, and the energy consumption evidence data includes spindle load current, feed axis servo current, cooling pump power and total vibration of key points. Analyze the functional satisfaction and system activation levels corresponding to the quality evidence data and the energy consumption evidence data, respectively. In the preset device status ring, query the target quadrant region corresponding to the functional satisfaction and the system activation. The device status ring includes the high-risk region in the upper left quadrant, the high-power region in the upper right quadrant, the steady-state region in the lower right quadrant, and the decay region in the lower left quadrant. Based on the target quadrant region, a visual remote anomaly warning is provided for the operating equipment.

2. The visual remote anomaly early warning method for precision manufacturing processes as described in claim 1, characterized in that, The analysis of the functional satisfaction and system activation levels corresponding to the quality evidence data and the energy consumption evidence data includes: The quality evidence data is classified according to rules to obtain the functional satisfaction level; The energy consumption evidence data is classified according to rules to obtain the system activation degree.

3. The visual remote anomaly early warning method for precision manufacturing processes as described in claim 2, characterized in that, The process of classifying the quality evidence data according to rules to obtain functional satisfaction includes: The servo tracking error, machining accuracy, and surface roughness signals in the quality evidence data are extracted by synchronous time windows to obtain the error sequence, dimensional deviation sequence, and noise feature sequence within the window; Calculate the maximum absolute value and standard deviation of the in-window error sequence; Calculate the center drift value of the size deviation sequence; Extract the high-frequency energy percentage of the noise feature sequence; The maximum absolute value, the standard deviation, the center drift value, and the high-frequency energy ratio are input into a preset quality rating table to query the functional satisfaction level in the quality rating table. The quality rating table refers to a two-dimensional lookup matrix with the maximum absolute value level as the row index, the standard deviation level as the column index, and the center drift coefficient and energy proportion coefficient as the downgrade criteria.

4. The visual remote anomaly early warning method for precision manufacturing processes as described in claim 2, characterized in that, The process of classifying the energy consumption evidence data according to rules to obtain the system activation degree includes: The spindle load current, feed axis servo current, cooling pump power, and total vibration at key points in the energy consumption evidence data are extracted by synchronous time windows to obtain current sequence, power sequence, and vibration sequence. Calculate the effective value of the current and the variance of the fluctuation in the current sequence; Calculate the mean and peak-to-peak values ​​of the power sequence; Calculate the total effective value of the vibration sequence; The effective value of the current, the fluctuation variance, the mean, the peak-to-peak value, and the total effective value are input into a preset energy consumption rating table to query the system activation level in the energy consumption rating table; The energy consumption rating table refers to a two-dimensional lookup matrix with effective value level as row index, fluctuation variance level as column index, and cooling power consumption coefficient and vibration total coefficient as upgrade standards.

5. The visual remote anomaly early warning method for precision manufacturing processes as described in claim 1, characterized in that, Before querying the target quadrant region corresponding to the functional satisfaction and system activation in the preset device status loop, the method further includes: A two-dimensional rectangular coordinate system is established with the functional pleasure level of historical periods as the horizontal axis and the system activation level of historical periods as the vertical axis. The positive and negative boundary value of the horizontal axis in the two-dimensional rectangular coordinate system is set as the functional pleasure threshold, and the positive and negative boundary value of the vertical axis in the two-dimensional rectangular coordinate system is set as the system activation threshold. Based on the functional pleasure threshold and the system activation threshold, the first quadrant, the second quadrant, the third quadrant and the fourth quadrant in the two-dimensional rectangular coordinate system are respectively regarded as the high-performance region of the upper right quadrant, the high-risk region of the upper left quadrant, the decay region of the lower left quadrant and the steady-state region of the lower right quadrant. The target quadrant region is determined by the two-dimensional rectangular coordinate system, the high-power region in the upper right quadrant, the high-risk region in the upper left quadrant, the decay region in the lower left quadrant, and the steady-state region in the lower right quadrant.

6. The visual remote anomaly early warning method for precision manufacturing processes as described in claim 1, characterized in that, The step of providing visualized remote anomaly warning for the operating equipment based on the target quadrant region includes: Render the device status loop on the remote monitoring interface to obtain the rendered status loop; The dynamic emotion points corresponding to functional pleasure and system activation in the target quadrant are displayed in real time within the rendering situation ring. By measuring the duration of the dynamic emotion point in the target quadrant, a high-risk anomaly warning for the operating equipment is generated, thereby enabling a visual remote anomaly warning for the operating equipment. By analyzing the movement trajectory of the dynamic emotion points between different target quadrants, a trend anomaly warning for the operating equipment is generated, thereby enabling visualized remote anomaly warning for the operating equipment.

7. The visual remote anomaly early warning method for precision manufacturing processes as described in claim 6, characterized in that, The step of generating an anomaly warning for the operating device by analyzing the movement trajectory of the dynamic emotion point between different target quadrants includes: When the dynamic emotion point switches from the steady-state area in the lower right quadrant to the decay area in the lower left quadrant, a performance degradation warning for the running device is generated. When the dynamic emotion point switches from the steady-state area in the lower right quadrant and the decay area in the lower left quadrant to the high-risk area in the upper left quadrant, an abnormal change warning is generated for the operating equipment. The performance degradation warning and the abnormal mutation warning are used to determine the trend anomaly warning of the operating equipment.

8. The visual remote anomaly early warning method for precision manufacturing processes as described in claim 1, characterized in that, After performing visual remote anomaly warning for the operating equipment based on the target quadrant region, the method further includes: Cluster analysis was performed on dynamic emotion points over a continuous period of time to obtain an emotion scatter plot; Calculate the first distribution area of ​​the steady-state region in the lower right quadrant of the emotion scatter plot; Calculate the second distribution area of ​​the high-risk area in the upper left quadrant and the decay area in the lower left quadrant of the emotion scatter plot; The ratio of the first distribution area to the second distribution area is used as the health decline index; Preventive maintenance recommendations for the operating equipment are generated using the health decline index.

9. The visual remote anomaly early warning method for precision manufacturing processes as described in claim 1, characterized in that, The collected quality evidence data and energy consumption evidence data of the operating equipment include: During each control cycle, the servo tracking error of the operating equipment is collected through the CNC servo interface; The machining accuracy of the operating equipment is obtained through the in-machine probe of the operating equipment. The surface roughness signal of the operating device is extracted through the microphone at the front end of the spindle of the operating device. The servo tracking error, the machining accuracy, and the surface roughness signal are used as quality evidence data. During each control cycle, the spindle load current and feed axis servo current of the operating device are collected through the driver current loop register; The cooling pump power of the operating equipment is collected using a bus power meter; The total vibration of key points of the operating equipment is collected using an accelerometer. The spindle load current, the feed axis servo current, the cooling pump power, and the total vibration at the key points are used as energy consumption evidence data.

10. A visual remote anomaly early warning system for precision manufacturing processes, characterized in that, The system includes: The data collection module is used to collect quality evidence data and energy consumption evidence data of the operating equipment during the precision manufacturing process. The quality evidence data includes servo tracking error, machining accuracy and surface roughness signals, and the energy consumption evidence data includes spindle load current, feed axis servo current, cooling pump power and total vibration of key points. The coordinate calculation module is used to analyze the functional satisfaction and system activation corresponding to the quality evidence data and the energy consumption evidence data, respectively. The quadrant query module is used to query the target quadrant region corresponding to the functional satisfaction and the system activation in a preset device status ring. The device status ring includes the high-risk region in the upper left quadrant, the high-power region in the upper right quadrant, the steady-state region in the lower right quadrant, and the decay region in the lower left quadrant. The anomaly warning module is used to provide visual remote anomaly warnings for the operating equipment based on the target quadrant region.

Images