A rolling mill encoder rotation speed deviation anti-interference dynamic monitoring and alarming method and device

Abstract

The present application relates to a kind of rolling mill encoder rotational speed deviation anti-interference dynamic monitoring alarm method and device, belong to the field of steel rolling automation monitoring;Method includes: real-time acquisition encoder rotational speed signal and rolling process stage identification signal;According to process stage identification signal, identify whether it is in the steel head tail transition period, if so, then enable dynamic time window filtering;In non head tail transition period, using adaptive dynamic deviation threshold for real-time comparison, the threshold is generated according to current steady-state rotational speed value and historical fluctuation variance Real-time calculation;When rotational speed deviation continuously exceeds dynamic deviation threshold to reach the preset confirmation length, output alarm signal, deviation regression normal after automatic reset.The present application is driven by the dynamic time window of working condition identification and the cooperative cooperation of adaptive threshold, fundamentally solves the contradiction between head tail interference and steady-state monitoring, realizes zero false alarm, zero false alarm, greatly reduces the demand of manual intervention.

Description

Technical Field

[0001] This invention relates to the field of automated monitoring technology for steel rolling mills, and in particular to a method and device for dynamic monitoring and alarm of speed deviation of rolling mill encoders to resist interference. Background Technology

[0002] The encoder of a rolling mill motor is a key sensor for speed closed-loop control and tension regulation. During actual rolling, the encoder signal experiences brief fluctuations due to factors such as sudden load changes and mechanical backlash during the bite and ejection stages. Existing monitoring systems often use fixed deviation thresholds and fixed filtering times, which easily generate false alarms during trial rolling of different specifications or acceleration / deceleration processes. This necessitates frequent manual resets, disrupting production and potentially masking actual faults. However, relaxing the threshold can lead to missed faults during steady-state operation, making it difficult to balance sensitivity and reliability.

[0003] The prior art CN103071685B discloses a vibration mark monitoring system and method for a 20-roll mill based on angular domain, which retrieves the corresponding parameter feature value from a lookup list as a threshold for comparison based on the actual speed of the mill. However, the threshold in this scheme is a pre-stored lookup value, which belongs to a static mapping relationship, rather than being dynamically calculated and generated based on the real-time fluctuation variance, and cannot adapt to the fluctuation differences under different operating conditions at the same speed.

[0004] In addition, the prior art CN122033040A discloses a method for determining the steel biting signal based on the tracking of the head and tail positions of each frame. It involves the determination of the steel biting and throwing signals, but it is only used for position tracking and looper control, and does not involve encoder speed deviation monitoring or dynamic threshold alarm.

[0005] In existing technologies, there are also schemes that use adaptive filtering for encoder signal processing, such as improving anti-interference performance through adaptive filtering. However, such schemes only involve speed measurement filtering itself and do not systematically integrate operating condition identification, dynamic window filtering, and adaptive threshold alarm.

[0006] In summary, none of the existing technologies have provided a complete technical solution that can simultaneously address the problem of resisting false alarms during periods of strong interference such as steel biting and throwing, and the problem of sensitive detection of real minor faults during steady-state operation. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide a method and device for dynamic monitoring and alarm of speed deviation of rolling mill encoder to achieve accurate monitoring with adaptive working conditions, eliminate false alarm interference and ensure that real faults are not missed.

[0008] The technical solution adopted by this invention to solve its technical problem is: a method for dynamic monitoring and alarm of speed deviation of rolling mill encoder, comprising the following steps: S1. Real-time acquisition of raw speed signal from the mill motor encoder and rolling process stage identification signal; S2. Identify whether the current period is the transition period between the head and tail of the steel bite according to the rolling process stage identification signal. If it is in this period, activate dynamic time window filtering, pause deviation comparison within the window or temporarily relax the deviation threshold to above the safety envelope. S3. During the non-head-to-tail transition period, an adaptive dynamic deviation threshold is used for real-time comparison. The dynamic deviation threshold is calculated and generated in real time based on the current steady-state speed value and the historical fluctuation variance. S4. When the speed deviation continuously exceeds the dynamic deviation threshold for a preset confirmation time, an alarm signal is output; when the deviation returns to the normal range, an automatic reset is performed.

[0009] Furthermore, the boundary of the dynamic time window in step S2 of the present invention is determined by multi-source signal fusion: the thermal detection signal is the main trigger, and the load current change rate dI / dt is the auxiliary verification; the bite-in stage window starts at the thermal detection ON moment and ends after the speed stabilizes and dI / dt is lower than the threshold; the throw-out stage window starts at the upstream rack thermal detection OFF moment and ends after the local rack speed stabilizes under no-load conditions.

[0010] Furthermore, the steel biting / spamming transition period confirmation mark described in this invention The calculation logic is as follows:

[0011] in, This indicates the status of the thermal detection signal. The real-time rate of change of load current. To verify the threshold for the rate of change of current.

[0012] Furthermore, the determination of rotational speed stability in this invention employs the sliding variance method, and rotational speed is determined to be stable when the following conditions are met:

[0013] Where M is the length of the stability evaluation window. The average rotational speed within the window. This is for the tolerance of speed fluctuation.

[0014] Furthermore, the adaptive dynamic deviation threshold mentioned in step S3 of the present invention The calculation formula is Where μ is the current steady-state speed average, σ is the standard deviation of the fluctuation over the past N sampling periods, and k is the confidence coefficient; μ and σ are calculated using the sliding window method, with a window length N of 200 sampling periods.

[0015] Furthermore, the k value described in this invention is dynamically determined based on the rolling specifications and steel grade: k=3.0 for steady-state rolling of ordinary carbon steel; k=3.5 for rolling of alloy steel or high-strength steel; k=4.0 for the trial rolling stage of the first steel after changing specifications, and the normal k value is restored from the second steel onwards; the adjustable range of the k value is 2.5~5.0.

[0016] Furthermore, the present invention addresses the adaptive dynamic deviation threshold. Apply a rate of change limit:

[0017] in, The maximum allowable rate of change is the threshold. The sampling period is specified; if the calculated value exceeds the limit, the output will gradually change according to the limit.

[0018] Furthermore, the present invention also includes a manual / automatic mode switching step: in automatic mode, all dynamic filtering and adaptive threshold logic are executed; in manual mode, the operator is allowed to set a fixed deviation threshold and a fixed filtering duration for parameter calibration during the trial rolling stage of new specifications; a built-in specification parameter table automatically loads the corresponding window parameters and threshold parameters according to the current rolling schedule number.

[0019] Furthermore, the present invention also includes an alarm classification processing step: Level 1 deviation exceeding the limit is only recorded without alarm; Level 2 deviation exceeding the limit triggers an audible and visual warning but does not interlock; Level 3 deviation exceeding the limit and the duration exceeding the safety limit triggers the mill deceleration or shutdown interlock.

[0020] Meanwhile, the present invention also provides a rolling mill encoder speed deviation anti-interference dynamic monitoring and alarm device for implementing the above method, including a signal interface unit for acquiring encoder speed signals and rolling process stage identification signals; a processor unit for performing working condition identification, dynamic time window filtering, adaptive threshold calculation and automatic reset logic; a human-machine interaction unit for mode switching, parameter setting and alarm information display; and an alarm output unit for outputting tiered audible and visual prompts and interlocking control signals.

[0021] The beneficial effect of this invention is that it solves the defects existing in the prior art. 1. By combining dynamic time windows driven by operating condition identification with adaptive thresholds, the contradiction between head-and-tail interference and steady-state monitoring is fundamentally resolved. The dynamic time window accurately covers the period of steel biting and throwing interference, avoiding the failure of fixed windows when speed changes or specifications are changed; the adaptive threshold automatically tightens or loosens according to the stability of the operating condition, ensuring that minor faults can be detected under steady state, while tolerating normal fluctuations without false alarms.

[0022] 2. The automatic reset mechanism eliminates the need for manual intervention and ensures production continuity. Monitoring automatically resumes once the deviation returns to the normal range.

[0023] 3. The tiered alarm strategy avoids false shutdowns caused by instantaneous spikes, significantly improving the availability and reliability of the monitoring system.

[0024] 4. The dynamic window boundary determination method for multi-source signal fusion (hot detection and dI / dt dual verification) ensures that the window accurately covers the interference period, with a window coverage rate of 99.6%, which is significantly better than the traditional fixed window scheme.

[0025] 5. The threshold change rate limiting strategy prevents sudden threshold changes from causing false alarms or missed alarms, ensuring a smooth threshold transition.

[0026] 6. Industrial operation tests have verified that, compared with the traditional fixed threshold method (47 false alarms / month) and fixed time window filtering method (18 false alarms / month), this invention achieves zero false alarms; the frequency of manual reset is reduced from 15 times per shift to 0 times; and the time for specification change debugging is shortened from 45 minutes to 10 minutes. Attached Figure Description

[0027] Figure 1 This is a logic flowchart of the monitoring and alarm method of the present invention; Figure 2 This is a schematic diagram of the hardware structure of the device of the present invention. Detailed Implementation

[0028] The present invention will now be described in further detail with reference to the accompanying drawings and preferred embodiments. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0029] like Figures 1-2 This invention discloses a method and device for dynamic monitoring and alarm of encoder speed deviation in a rolling mill. This embodiment is applied to the motor monitoring of a second-line rolling mill in a bar mill. The processor unit acquires encoder speed ω and process stage identification signals such as thermal detection and load current at 5ms cycles.

[0030] The method includes the following steps: S1. Real-time acquisition of raw speed signal from the mill motor encoder and rolling process stage identification signal; S2. Identify whether the current period is the transition period between the head and tail of the steel bite according to the rolling process stage identification signal. If it is in this period, activate dynamic time window filtering, pause deviation comparison within the window or temporarily relax the deviation threshold to above the safety envelope. S3. During the non-head-to-tail transition period, an adaptive dynamic deviation threshold is used for real-time comparison. The dynamic deviation threshold is calculated and generated in real time based on the current steady-state speed value and the historical fluctuation variance. S4. When the speed deviation continuously exceeds the dynamic deviation threshold for a preset confirmation time, an alarm signal is output; when the deviation returns to the normal range, an automatic reset is performed.

[0031] Specifically: I. System Initialization and Data Acquisition The processor unit of this invention acquires the raw speed signal ω fed back by the rolling mill motor encoder in real time through the signal interface unit at a period of 5ms, as well as key process stage identification signals such as thermal inspection signals and motor load current signals from the production line. These multi-source signals constitute the data foundation for all subsequent dynamic judgments and calculations.

[0032] II. Dynamic Time Window Filtering Based on Operating Condition Identification

[0033] After obtaining real-time data, the system immediately executes the core operating condition identification logic. This step is performed because it was found that during the transition phases at the beginning and end of the steel biting and throwing stages, drastic changes in load can cause instantaneous fluctuations in rotational speed. If conventional monitoring is used at this time, it is very easy to cause false alarms.

[0034] Therefore, in step S2, the system accurately determines whether it is currently in the transition period between steel biting and steel throwing based on process stage identification signals such as thermal detection and load current. Once this period is identified, the system immediately activates the dynamic time window filtering mechanism. Within this window, the system will pause deviation comparison or temporarily relax the deviation threshold to above a safe envelope, thereby actively avoiding known, periodic operating condition interference.

[0035] Furthermore, to ensure the accuracy of the window boundaries, this embodiment employs a multi-source signal fusion method to determine the start and end of the window.

[0036] Specifically, the confirmation criteria for the steel biting / spamming transition period. From thermal detection signal With load current change rate The joint decision-making process, with the following logical relationship:

[0037] in: This indicates the status of the thermal detection signal (1=ON / steel detected, 0=OFF). The real-time rate of change of the load current is calculated using a first-order difference: , The sampling period is 5ms. The current change rate verification threshold is used to eliminate false signals caused by thermal detection mis-triggers or mechanical vibrations.

[0038] This auxiliary verification condition can effectively eliminate false signals caused by false triggering of thermal detection signals or mechanical vibration, ensuring that the dynamic window is only opened during the interference period when a real load change occurs.

[0039] Based on the above fusion determination, the bite-in phase window begins when the thermal detection signal turns ON and ends when the rotational speed reaches a stable state. After falling below the preset threshold, the ejection phase window begins when the upstream rack thermal detection signal turns OFF and ends after the rack speed reaches a stable no-load state.

[0040] Here, the steady-state of rotational speed is quantitatively determined using the sliding variance method, and the criteria are as follows:

[0041] The stability evaluation window length is set to 20 sampling periods (i.e., 100ms). The average rotational speed within the window; To account for the speed fluctuation tolerance, the base value is set to 0.3% of the rated speed.

[0042] This dynamic adaptation mechanism ensures that the window can accurately cover the actual interference period, effectively avoiding the problem of traditional fixed windows failing when changing speed or specifications.

[0043] III. Generation and Optimization of Adaptive Dynamic Deviation Threshold

[0044] When step S2 identifies a steady-state period that is not in a transitional phase, the system seamlessly switches to a more sensitive monitoring mode. In this mode, in order to capture real minor faults while tolerating normal operating condition fluctuations, this invention employs an adaptive dynamic deviation threshold for real-time comparison.

[0045] In step S3, the dynamic deviation threshold It is not a fixed constant, but rather calculated in real time based on the current steady-state speed value and historical fluctuation variance. The calculation formula is as follows:

[0046] Where μ is the current steady-state average speed, σ is the standard deviation of the fluctuation over the past N sampling periods, and k is the confidence coefficient. The advantage of this calculation method is that the threshold will automatically tighten or loosen depending on the stability of the operating conditions: the threshold will be loosened when the fluctuation is large to prevent false alarms, and tightened when the fluctuation is small to improve sensitivity.

[0047] To ensure the accuracy and adaptability of this threshold calculation, this embodiment also provides the following specific implementation details: Statistical parameter calculation method: The mean μ and standard deviation σ are calculated using the sliding window method, with a window length N of 200 sampling periods (i.e., a 1-second data window). To avoid outliers contaminating the statistics, outliers are removed from the window data using the 3σ criterion before calculation.

[0048] The confidence coefficient k has a dynamic value: k is not a fixed constant, but a dynamic parameter related to the rolling specifications and steel grade. For example, k=3.0 for steady-state rolling of ordinary carbon steel; k=3.5 for rolling alloy steel or high-strength steel (where load fluctuations are large); and k=4.0 during the trial rolling stage of the first steel after changing specifications (relaxing the threshold to avoid false alarms). The k value automatically returns to the normal value starting with the second steel. The k value can also be manually corrected via HMI, with the correction range limited to 2.5~5.0.

[0049] Threshold change rate limiting strategy: To prevent false alarms or missed alarms caused by large jumps in threshold due to sudden changes in operating conditions, this invention limits the threshold change rate. A rate of change constraint was imposed, and the constraint condition is as follows: ,in, The maximum allowable rate of change of the threshold is set to 2% / s of the rated speed in this embodiment. If the calculated value exceeds this limit, the output will be gradually changed according to the limit value, thereby ensuring a smooth transition of the threshold.

[0050] IV. Alarm Detection, Classification and Automatic Reset

[0051] After the dynamic deviation threshold is determined, the system executes the final alarm logic judgment. In step S4, when the real-time speed deviation Δω continuously exceeds the dynamic deviation threshold... The system will only output an alarm signal when the preset confirmation time (e.g., 200ms) is reached, which effectively avoids interference from instantaneous spike pulses.

[0052] Furthermore, to differentiate between anomalies of varying severity, this invention introduces a graded alarm processing mechanism. Under this mechanism, a Level 1 deviation exceeding the limit is recorded without triggering an alarm; a Level 2 deviation exceeding the limit triggers an audible and visual warning but does not cause interlocking; only a Level 3 deviation exceeding the limit and lasting longer than the safety limit triggers a mill deceleration or shutdown interlock. This graded strategy effectively prevents accidental shutdowns caused by momentary spike interference, ensuring both safety and production efficiency.

[0053] Once the deviation returns to the normal range (for example, Δω falls back to...), If the false alarm occurs within 500ms, the system will automatically reset, restoring the monitoring status without manual intervention. This mechanism completely eliminates the adverse effects caused by frequent false alarms in traditional solutions, greatly reducing the workload of operators.

[0054] V. Auxiliary Functions: Mode Switching, Parameter Configuration, and Data Recording

[0055] In addition to the core monitoring logic described above, this invention also provides a variety of auxiliary functions to enhance its adaptability to industrial sites.

[0056] The system supports manual / automatic mode switching to meet the needs of different scenarios. In automatic mode, the system executes all the dynamic filtering and adaptive threshold logic mentioned above; while in manual mode, the operator can set a fixed deviation threshold and a fixed filtering duration through the human-machine interface for parameter calibration and verification during the trial rolling stage of new specifications.

[0057] In addition, the system has a built-in specification parameter table, which can automatically load the corresponding window parameters and threshold parameters based on the current rolling schedule number. An example configuration is shown in the table below:

[0058] During trial rolling of new specifications, operators can manually adjust the above parameters online and save them to the formula library, which greatly shortens the debugging time for changing specifications.

[0059] Finally, the system also features historical data recording and analysis capabilities, capable of recording the original waveform segments, process parameters at the time, dynamic threshold values, and filter window status for each deviation exceeding the limit event. This data not only supports post-event playback analysis to distinguish whether the fault source is encoder body failure, mechanical transmission backlash, or electrical interference, but also provides valuable training samples for subsequent iterative optimization of the dynamic threshold algorithm.

[0060] VI. Industrial Operation Testing and Effect Verification

[0061] To verify the practical effectiveness of this invention, a three-month industrial operation test was conducted on a second-line rolling mill in a bar mill. The test results were compared with those of traditional technical solutions, and the statistical data are as follows:

[0062] The test results fully demonstrate that the present invention, through the coordinated use of dynamic windows driven by operating condition recognition and adaptive thresholds, completely eliminates false alarms while ensuring zero missed reports of real faults, significantly reduces the need for manual intervention, and greatly shortens the debugging cycle of new specifications, achieving a technical effect that combines sensitivity and reliability.

[0063] The above description is only a specific embodiment of the present invention. Various examples and illustrations do not constitute a limitation on the substantive content of the present invention. Those skilled in the art can make modifications or variations to the above-described specific embodiments after reading the specification without departing from the substance and scope of the invention.

Claims

1. A method for dynamic monitoring and alarming of speed deviation of a rolling mill encoder, characterized in that: Includes the following steps: S1. Real-time acquisition of raw speed signal from the mill motor encoder and rolling process stage identification signal; S2. Identify whether the current period is the transition period between the head and tail of the steel bite according to the rolling process stage identification signal. If it is in this period, activate dynamic time window filtering, pause deviation comparison within the window or temporarily relax the deviation threshold to above the safety envelope. S3. During the non-head-to-tail transition period, an adaptive dynamic deviation threshold is used for real-time comparison. The dynamic deviation threshold is calculated and generated in real time based on the current steady-state speed value and the historical fluctuation variance. S4. When the speed deviation continuously exceeds the dynamic deviation threshold and reaches the preset confirmation time, an alarm signal is output. Once the deviation returns to the normal range, an automatic reset is executed.

2. The method for dynamic monitoring and alarming of speed deviation of a rolling mill encoder as described in claim 1, characterized in that: The boundary of the dynamic time window in step S2 is determined by multi-source signal fusion: the thermal detection signal is the main trigger, and the load current change rate dI / dt is the auxiliary verification; the bite-in stage window starts at the thermal detection ON moment and ends after the rotation speed stabilizes and dI / dt is below the threshold. The throw phase window begins when the upstream rack is OFF during hot inspection and ends after the rack speed stabilizes under no-load conditions.

3. The method for dynamic monitoring and alarming of speed deviation of a rolling mill encoder as described in claim 2, characterized in that: The steel biting / spamming transition period confirmation mark The calculation logic is as follows: ， in, This indicates the status of the thermal detection signal. The real-time rate of change of load current. To verify the threshold for the rate of change of current.

4. The method for dynamic monitoring and alarming of speed deviation of a rolling mill encoder as described in claim 2, characterized in that: The determination of rotational speed stability uses the sliding variance method. Rotational speed is considered stable when the following conditions are met: Where M is the length of the stability evaluation window, The average rotational speed within the window. This is for the tolerance of speed fluctuation.

5. The method for dynamic monitoring and alarming of speed deviation of a rolling mill encoder as described in claim 1, characterized in that: The adaptive dynamic deviation threshold mentioned in step S3 The calculation formula is Where μ is the current steady-state speed average, σ is the standard deviation of the fluctuation over the past N sampling periods, and k is the confidence coefficient; μ and σ are calculated using the sliding window method, with a window length N of 200 sampling periods.

6. The method for dynamic monitoring and alarming of speed deviation of a rolling mill encoder as described in claim 5, characterized in that: The k value is dynamically determined based on the rolling specifications and steel grade: k=3.0 for steady-state rolling of ordinary carbon steel; k=3.5 for rolling of alloy steel or high-strength steel; k=4.0 for the trial rolling stage of the first steel after changing specifications, and the normal k value is restored from the second steel; the adjustable range of the k value is 2.5~5.

0.

7. The method for dynamic monitoring and alarming of speed deviation of a rolling mill encoder as described in claim 5, characterized in that: For the adaptive dynamic deviation threshold Apply a rate of change limit: ， in, The maximum allowable rate of change is the threshold. The sampling period is specified; if the calculated value exceeds the limit, the output will gradually change according to the limit.

8. The method for dynamic monitoring and alarming of speed deviation of a rolling mill encoder as described in claim 1, characterized in that: It also includes a manual / automatic mode switching step: in automatic mode, all dynamic filtering and adaptive threshold logic are executed; in manual mode, the operator can set a fixed deviation threshold and a fixed filtering duration for parameter calibration during the trial rolling stage of new specifications; a built-in specification parameter table automatically loads the corresponding window parameters and threshold parameters according to the current rolling schedule number.

9. The method for dynamic monitoring and alarming of speed deviation of a rolling mill encoder as described in claim 1, characterized in that: It also includes alarm classification and processing steps: Level 1 deviation exceeding the limit is only recorded without alarm; Level 2 deviation exceeding the limit triggers audible and visual prompts but does not interlock; Level 3 deviation exceeding the limit and the duration exceeding the safety limit triggers mill deceleration or shutdown interlock.

10. A rolling mill encoder speed deviation anti-interference dynamic monitoring and alarm device, used to implement the method described in any one of claims 1-9, characterized in that: It includes a signal interface unit for acquiring encoder speed signals and rolling process stage identification signals; a processor unit for performing working condition identification, dynamic time window filtering, adaptive threshold calculation, and automatic reset logic; a human-machine interaction unit for mode switching, parameter setting, and alarm information display; and an alarm output unit for outputting tiered audible and visual prompts and interlocking control signals.

Images