Triaxial acceleration sensor adaptive monitoring method, device, equipment and medium

Abstract

The application discloses a kind of triaxial acceleration sensor adaptive monitoring method, device, equipment and medium.It includes: obtaining the time domain triaxial vibration voltage signal collected by sensor;When in stable operation condition, the total effective value electric signal energy of triaxial vibration voltage signal in preset time window is calculated, and voltage sensitivity degradation coefficient is calculated;According to voltage sensitivity degradation coefficient and preset physical vibration safety limit value, calculate reference voltage alarm threshold;According to the acceleration data of sensor, calculate first-order difference rate and second-order difference rate, determine whether it is external transient clutter interference now;When it is determined that external transient clutter interference, calculate initial escape lift, build escape attenuation model for calculating dynamic voltage alarm threshold;The sample proportion of triaxial vibration voltage signal exceeding dynamic voltage alarm threshold in short-time fast window and long-time slow window is counted, and the state interval to which it belongs now is set.The accuracy and reliability of triaxial acceleration sensor monitoring can be improved.

Description

Technical Field

[0001] This invention belongs to the field of acceleration sensor signal processing technology, specifically relating to an adaptive monitoring method, device, equipment, and medium for a triaxial acceleration sensor. Background Technology

[0002] In industrial predictive maintenance systems for rotating machinery such as motors, fans, and pumps, triaxial accelerometers are widely deployed on the surface of the motor housing to collect mechanical vibration signals in real time and output effective values ​​and other characteristics, thereby determining whether an alarm exceeds the standard. Due to its convenient deployment and real-time response, this method has become one of the mainstream technologies for industrial equipment condition monitoring.

[0003] However, the harsh industrial environment and complex interference factors mean that the above-mentioned alarm judgment system based on fixed thresholds still faces the following technical bottlenecks in actual operation, which restrict the accuracy of alarms and the long-term reliability of the system:

[0004] I. Risk of missed detection caused by sensor voltage sensitivity degradation

[0005] Under conditions of high temperature, high humidity, or prolonged exposure to severe high-frequency vibration, the internal MEMS sensitive structure or piezoelectric element of a triaxial accelerometer undergoes irreversible physical aging, causing the sensor's voltage sensitivity to monotonically decrease over time. Since existing systems typically use the factory-calibrated sensitivity coefficient as the basis for signal conversion, when the actual sensitivity decreases, the electrical signal output corresponding to the same physical vibration amplitude will significantly decrease. This results in the calculated physical vibration quantities such as acceleration, velocity, or displacement being significantly lower than the actual values. This deviation directly leads to the system's inability to identify gradually deteriorating vibration conditions, resulting in missed detections of early equipment failures and severe degradation.

[0006] II. Frequent false alarms caused by transient unstructured clutter

[0007] Industrial environments are characterized by complex electromagnetic environments, often accompanied by transient interference signals such as random electromagnetic surges, inverter switching noise, nearby impacts, or unsteady mechanical noise. These transient spike signals are characterized by large amplitude, short duration, and concentrated energy in the time domain. Their first-order amplitude can easily exceed the system's alarm limits instantaneously. Due to the lack of ability to distinguish between structured and unstructured signal sources, it is difficult to effectively differentiate between genuine mechanical fault impacts and external transient noise, leading to frequent false alarms.

[0008] III. The contradiction between latency and resource overhead in state debouncing algorithms at the edge.

[0009] To suppress false alarms, traditional solutions often employ long sliding windows to smooth and filter feature values ​​at the central monitoring platform or edge devices. While this method can reduce the impact of transient fluctuations to some extent, it has two significant drawbacks: First, long-window smoothing requires edge sensor nodes or low-power lower-level machines to cache a large amount of historical waveform data, significantly increasing memory usage and computational overhead, placing a heavy burden on resource-constrained edge devices. Second, while eliminating noise, smoothing filtering also dilutes and delays the characteristic responses of truly sudden early severe faults, causing the system to lag in identifying malicious faults and responding to shutdown protection, thus facing higher equipment failure and security risks.

[0010] Therefore, the accuracy and reliability of triaxial accelerometer monitoring need to be improved. Summary of the Invention

[0011] The purpose of this invention is to provide an adaptive monitoring method, device, equipment, and storage medium for a triaxial accelerometer, which can improve the accuracy and reliability of triaxial accelerometer monitoring.

[0012] The first aspect of this invention discloses an adaptive monitoring method for a triaxial accelerometer, comprising:

[0013] Acquire the time-domain triaxial vibration voltage signal collected by the sensor;

[0014] When operating under stable conditions, calculate the total effective value of the triaxial vibration voltage signal energy within a preset time window, and calculate the voltage sensitivity degradation coefficient based on the total effective value of the voltage signal energy.

[0015] The preset physical vibration safety limit is adjusted in real time and closed-loop adaptively based on the voltage sensitivity degradation coefficient to obtain the reference voltage alarm threshold.

[0016] Calculate the absolute values ​​of the first-order differential rate of change and the second-order differential rate of change based on the acceleration data from the sensor. Then, based on the first-order differential rate of change, the second-order differential rate of change, and the physical inertial response limit, determine whether the current situation is due to external transient clutter interference.

[0017] When the external transient clutter interference is determined, the initial escape rise is calculated, and an escape attenuation model for calculating the dynamic voltage alarm threshold is constructed based on the initial escape rise.

[0018] The proportion of samples in which the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold within the short-time fast window and the long-time slow window is statistically analyzed, and the current state interval is set according to the sample proportion.

[0019] In some embodiments, the proportion of samples in which the vibration voltage signal exceeds the dynamic voltage alarm threshold within a short-time fast window and a long-time slow window is statistically analyzed, and the current state interval is set according to the sample proportion, including:

[0020] The proportion of samples in which the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold within a short-time window is statistically analyzed to obtain the first proportion;

[0021] The proportion of samples in which the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold within a long-term slow window is statistically analyzed to obtain the second proportion;

[0022] When the first percentage is greater than the first threshold, the current state interval is set as the attention zone; when the state interval is the attention zone, when the second percentage is greater than the second threshold and the duration exceeds the preset duration, or when the first percentage is greater than the third threshold, the current state interval is set as the abnormal zone.

[0023] In some embodiments, the proportion of samples in a target window whose triaxial vibration voltage signals exceed the dynamic voltage alarm threshold is statistically analyzed to obtain a target proportion. The target window is a short-time fast window or a long-time slow window, and the target proportion is a first proportion or a second proportion, including:

[0024] Based on whether the triaxial vibration voltage signal at each sampling moment within the target window exceeds the dynamic voltage alarm threshold at that sampling moment, a flag corresponding to each sampling moment is set to obtain a flag queue;

[0025] The target percentage is obtained by performing a moving average on the flag queue.

[0026] In some embodiments, the expression for calculating the total effective value of the triaxial vibration voltage signal energy within a preset time window is as follows:

[0027]

[0028] in, The total effective value of the electrical signal energy. This represents the total number of voltage signals within a preset time window. , , For the first A triaxial vibration voltage signal;

[0029] The expression for calculating the voltage sensitivity degradation coefficient based on the total effective value of the electrical signal energy is as follows:

[0030]

[0031] in, for Voltage sensitivity degradation coefficient at time t. The smoothing time constant has a value ranging from 0.95 to 0.999. This refers to the sensor's factory-specified sensitivity. These are the reference physical constants for steady-state vibration.

[0032] In some embodiments, a triaxial accelerometer is mounted on the surface of the motor housing to determine whether the operating condition is stable, including:

[0033] The motor is monitored in real time to obtain the average active power and average speed. The active power fluctuation rate is calculated based on the average active power and the speed fluctuation rate is calculated based on the average speed.

[0034] When the active power fluctuation rate is lower than the first prior threshold and the speed fluctuation rate is lower than the second prior threshold, it is determined to be in a stable operating condition.

[0035] In some embodiments, the escape decay model is expressed as follows:

[0036] in, This is the initial escape lift. The reference voltage alarm threshold, To determine the moment when external transient clutter interference is triggered, The escape decay time constant has a value range of 1ms to 1000ms.

[0037] A second aspect of this invention discloses an adaptive monitoring device for a triaxial accelerometer, comprising:

[0038] A multi-channel signal acquisition module is used to acquire time-domain triaxial vibration voltage signals collected by the sensor;

[0039] The voltage sensitivity degradation estimation module is used to calculate the total effective value electrical signal energy of the triaxial vibration voltage signal within a preset time window when the operating conditions are stable, and to calculate the voltage sensitivity degradation coefficient based on the total effective value electrical signal energy.

[0040] The reference threshold correction module is used to perform closed-loop adaptive real-time adjustment of the preset physical vibration safety limit based on the voltage sensitivity degradation coefficient to obtain the reference voltage alarm threshold.

[0041] The transient clutter interference determination module is used to calculate the first-order differential rate of change and the second-order differential rate of change based on the acceleration data of the sensor, and to determine whether there is external transient clutter interference based on the absolute value of the first-order differential rate of change, the absolute value of the second-order differential rate of change and the physical inertial response limit.

[0042] The dynamic threshold escape module is used to calculate the initial escape rise when it is determined to be external transient clutter interference, and to construct an escape attenuation model for calculating the dynamic voltage alarm threshold based on the initial escape rise.

[0043] The cascaded de-jitter state machine module is used to count the proportion of samples in the short-time fast window and the long-time slow window where the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold, and to set the current state interval based on the sample proportion.

[0044] In some embodiments, the cascaded de-jitter state machine module includes a percentage calculation unit and a determination unit. The percentage calculation unit is used to calculate the percentage of samples in a short-term fast window where the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold to obtain a first percentage, and to calculate the percentage of samples in a long-term slow window where the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold to obtain a second percentage. The determination unit is used to set the current state interval as a region of interest when the first percentage > a first threshold; when the state interval is a region of interest, if the second percentage > a second threshold and the duration exceeds a preset duration, or if the first percentage > a third threshold, the current state interval is set as an abnormal region.

[0045] A third aspect of the present invention discloses an electronic device, including a memory storing executable program code and a processor coupled to the memory; the processor calls the executable program code stored in the memory to execute the triaxial accelerometer adaptive monitoring method disclosed in the first aspect.

[0046] The fourth aspect of the present invention discloses a computer-readable storage medium storing a computer program, wherein the computer program causes a computer to execute the triaxial accelerometer adaptive monitoring method disclosed in the first aspect.

[0047] The beneficial effects of this invention are as follows: by extracting the total effective value of the vibration voltage signal energy within the steady-state operating range, online automatic evaluation and compensation of the sensor voltage sensitivity are achieved, preventing false alarms caused by performance degradation of sensitive components. By introducing absolute value criteria for first- and second-order differential rates of change that match mechanical and physical rigidity limits, non-physical transient clutter is identified. Utilizing an escape decay model, the judgment threshold can be adaptively raised during interference and recovered exponentially, suppressing the probability of false alarms caused by transient noise. By statistically analyzing the over-limit ratio of short-time fast windows and long-time slow windows, true sudden early severe faults are identified. This improves the accuracy and reliability of triaxial accelerometer monitoring. Attached Figure Description

[0048] The accompanying drawings illustrate specific examples of the technical solutions described in this invention and, together with the detailed embodiments, form part of the specification, serving to explain the technical solutions, principles, and effects of this invention.

[0049] Unless otherwise specified or defined, the same reference numerals in different figures represent the same or similar technical features, and different reference numerals may be used to represent the same or similar technical features.

[0050] Figure 1 This is a flowchart of an adaptive monitoring method for a triaxial accelerometer according to an embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of the dynamic escape threshold according to an embodiment of the present invention;

[0052] Figure 3 This is a schematic diagram of the structure of the triaxial accelerometer adaptive monitoring device according to an embodiment of the present invention;

[0053] Figure 4 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0054] Unless otherwise specified or defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. When combined with the technical solutions of the invention in a real-world scenario, all technical and scientific terms used herein may also have meanings corresponding to the purpose of achieving the technical solutions of the invention. The terms "first," "second," etc., used herein are merely for distinguishing names and do not represent a specific number or order. The term "and / or," as used herein, includes any and all combinations of one or more of the associated listed items.

[0055] Unless otherwise specified or defined, the terms "described" or "the" as used herein refer to the technical features or technical content mentioned or described prior to the relevant section, which may be the same as or similar to the technical features or technical content mentioned herein. Furthermore, the terms "comprising" and "having," and any variations thereof, as used herein, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such processes, methods, products, or apparatus.

[0056] This invention discloses an adaptive monitoring method using a triaxial accelerometer. The accelerometer is mounted on the surface of a motor housing and includes a triaxial accelerometer and a microprocessor unit (MCU). This method is used to improve the accuracy and reliability of motor condition monitoring results during adaptive monitoring. The method can be implemented via computer programming and runs in the firmware of the microprocessor unit. The execution entity can also be an electronic device such as a computer, laptop, or tablet, or a control chip embedded in an electronic device; this invention does not limit this. It should be noted that the adaptive monitoring method of this invention is not limited to motor condition monitoring; its application scope can be widely extended to all mechanical equipment with rotating, reciprocating, or linear motion components, including but not limited to: fans, compressors, pumps, gearboxes, centrifuges, generators, turbomachinery, machine tool spindles, conveying equipment, and large slewing bearing structures. In any application scenario that uses a triaxial accelerometer to collect vibration signals and perform condition judgment, this invention can be used to achieve sensitivity drift compensation, transient interference suppression, and low-overhead edge computing, thereby improving the accuracy of monitoring results and the reliability of alarms.

[0057] To facilitate understanding of the present invention, specific embodiments of the present invention will be described in more detail below with reference to the accompanying drawings.

[0058] like Figure 1 As shown, the specific steps include:

[0059] Step S100: Acquire the time-domain triaxial vibration voltage signal collected by the sensor;

[0060] A triaxial accelerometer deployed on the surface of the motor housing is used to collect vibration voltage signals in the three orthogonal directions of X, Y, and Z in real time, which are denoted as follows: , , .

[0061] Step S200: When operating under stable conditions, calculate the total effective value of the triaxial vibration voltage signal energy within a preset time window, and calculate the voltage sensitivity degradation coefficient based on the total effective value of the voltage signal energy.

[0062] The method for determining stable operating conditions is set according to the application scenario of the triaxial accelerometer. In this embodiment, the triaxial accelerometer is mounted on the surface of the motor housing. The process of determining whether it is in a stable operating condition includes: real-time acquisition of the motor's active power and speed parameters from the motor power supply control box, and calculation of the average active power. and average speed Based on the average active power and the standard deviation of active power Calculate active power fluctuation rate Based on the average speed and the standard deviation of the speed Calculate the rotational speed fluctuation rate When the active power fluctuation rate is lower than the first prior threshold And the rotational speed fluctuation rate is lower than the second prior threshold. When this condition is met, the motor is determined to be in a stable operating condition. The specific expression is:

[0063]

[0064] Furthermore, to effectively address various normal operating conditions of the motor (such as no-load, light-load, and full-load variable operating conditions), several operating condition subspaces can be partitioned based on historical multi-dimensional operating characteristics (including speed, current, and total vibration) using unsupervised clustering algorithms (such as principal component analysis (PCA) combined with the noisy density-based clustering method DBSCAN). For each operating condition subspace, its historical FIFO circular historical cache is maintained independently, and the local threshold for this operating condition is calculated separately. When a change in operating condition is detected, a smooth loading is implemented to prevent threshold invalidation caused by normal switching of operating conditions.

[0065] For example, steady state can also be determined and locked by stator current sideband modulation MCSA or active power periodic work characteristics.

[0066] Under stable operating conditions, theoretically, the total effective value of the electrical signal energy of the motor's triaxial physical vibration should stabilize at a reference physical constant that matches the current operating conditions. (The unit is g. This reference physical constant can be based on the initial stable statistical value of the same type of equipment or the rated physical design limit of the motor and calibrated online.) Therefore, the voltage sensitivity degradation coefficient can be calculated based on the total effective value of the electrical signal energy and the reference physical constant.

[0067] Specifically, the triaxial vibration voltage signal is calculated within a preset time window. Total effective value of electrical signal energy within The expression is:

[0068]

[0069] in, The total effective value of the electrical signal energy. Preset time window The total number of voltage signals inside, , , For the first A triaxial vibration voltage signal. Preset time window. This refers to a time window of a preset length, with its start and end points being sliding.

[0070] Combined with the sensor's factory-rated sensitivity and steady-state vibration reference physical constants Calculate the voltage sensitivity degradation coefficient The specific expression is:

[0071]

[0072] in, for Voltage sensitivity degradation coefficient at time; The smoothing time constant has a value ranging from 0.95 to 0.999 to filter out short-term random disturbances. This refers to the sensor's factory-specified sensitivity. These are the reference physical constants for steady-state vibration. =1.0.

[0073] Step S300: Based on the voltage sensitivity degradation coefficient, perform closed-loop adaptive real-time adjustment of the preset physical vibration safety limit to obtain the reference voltage alarm threshold;

[0074] Based on voltage sensitivity degradation coefficient and preset physical vibration safety limits Closed-loop adaptive real-time adjustment is performed to obtain the reference voltage alarm threshold. The specific expression is:

[0075]

[0076] when Voltage alarm threshold when aging occurs Closed-loop adjustment enables online automatic evaluation and compensation of sensor voltage sensitivity, preventing delayed alarms or missed alarms caused by performance degradation of sensitive components, and reducing the probability of missed alarms due to aging.

[0077] Step S400: Calculate the first-order differential rate of change and the second-order differential rate of change based on the acceleration data from the sensor. Based on the first-order differential rate of change, the second-order differential rate of change, and the physical inertial response limit, determine whether the current situation is due to external transient clutter interference.

[0078] Using the upper limit of the physical structure inertial response of the motor as a benchmark, if the first and second time-domain differential change rates of acceleration exceed the inertial limit of the motor, it means that the current spike signal exceeds the physical inertial limit of the mechanical rigid structure of the motor. Therefore, it is determined that it has encountered external random pulse noise, thus distinguishing between instantaneous and sudden non-structural spike electrical signals and normal equipment physical vibration.

[0079] Specifically, the instantaneous first-order differential rate of change of acceleration is calculated based on the acceleration data from the sensor. and second-order rate of change ,in Assume the physical inertial response limit is and The logic for determining whether the current disturbance is caused by external transient clutter is as follows:

[0080]

[0081] That is, if any of the above conditions are met, the measured signal is determined to exceed the upper limit of the possibility of physical vibration of the motor and belongs to non-mechanical pulse noise interference (such as electromagnetic surge, random knocking, etc.).

[0082] Step S500: When it is determined to be external transient clutter interference, calculate the initial escape rise and construct an escape attenuation model for calculating the dynamic voltage alarm threshold based on the initial escape rise.

[0083] At the instant external transient clutter interference is detected, i.e., when external non-mechanical pulse interference is detected, the initial escape rise is calculated based on the peak amplitude. An escape attenuation model is constructed to slide down and align to the reference voltage alarm threshold in an exponential decay manner. The dynamic voltage alarm threshold is then calculated in real time based on the escape attenuation model. This causes a step change in the judgment threshold to suppress noise. If the threshold is not exceeded, it is judged as normal physical vibration, and the reference voltage alarm threshold is maintained. .

[0084] Specifically, the expression for the escape decay model is:

[0085]

[0086] in, This is the initial escape lift. The reference voltage alarm threshold, To determine the moment when external transient clutter interference is triggered, The escape decay time constant has a specific value range of 1ms to 1000ms, preferably 10ms to 100ms. The initial escape lift is calculated as follows: or,

[0087] .in, The magnification factor has a specific value range of 1.0 to 10.0, preferably 1.2 to 3.0; The first-order differential rate of change of the vibration voltage signal. It represents the second-order differential rate of change of the vibration voltage signal.

[0088] like Figure 2As shown, when the first / second-order differential rate of change exceeds the physical upper limit, the escape threshold is immediately activated, raising the judgment threshold and causing it to decay exponentially over time to the reference voltage alarm threshold. This can suppress the probability of false alarms caused by transient noise.

[0089] Step S600: Calculate the proportion of samples in the short-time fast window and long-time slow window where the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold, and set the current state interval based on the sample proportion.

[0090] Short-time fast windows (e.g., 16 sampling points) are used to respond to sudden events, enabling rapid triggering and effective capture of malignant mutation states; long-time slow windows (e.g., 256 sampling points) are used to smooth out normal degradation trends, effectively suppressing false alarms caused by occasional environmental fluctuations. Both short-time fast windows and long-time slow windows are sliding windows.

[0091] The proportion of samples in a short-term fast window where the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold is calculated to obtain the first proportion; the proportion of samples in a long-term slow window where the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold is calculated to obtain the second proportion; if the first proportion is higher than the first threshold, the current state interval is set as the attention zone; in the attention zone, if the second proportion is higher than the second threshold and the duration exceeds the preset duration, or if the first proportion is higher than the third threshold, the current state interval is set as the abnormal zone, and the corresponding equipment warning signal or safety shutdown control signal is output to prevent the occurrence of serious motor accidents.

[0092] Specifically, short-time windows are statistically analyzed on the lower-level machine side. (length is) ) and long-term slow window (length is) ,and The proportion of samples with excessive time-domain voltage within a given time period. Taking the first proportion of short-time windows as an example, the calculation process is as follows:

[0093] Within each sampling time k within a short time window, determine whether the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold at the sampling time, and set a binary flag based on the determination result:

[0094] ,in, Let this be the indicator function. The binary flags from all sampling times form a flag queue. The first proportion is the moving average of this flag queue, calculated using the following formula:

[0095]

[0096] After the above operations, not only can the problem of secondary overshoot caused by threshold decay be completely eliminated, but the sample ratio calculation can also be performed directly using the cascaded state determination layer. There is no need to cache massive amounts of high-frequency raw waveform data in memory. Each slide only requires simple bit shifting and accumulation, which reduces the memory overhead (only 2 bytes are needed for 16 sampling points) and computational overhead at the edge by several orders of magnitude.

[0097] when At that time, it moves from the normal zone to the attention zone; in the attention zone, if Continuously exceeding the prescribed time, or meeting the following conditions When an abnormal / alarm zone is reached, the system will directly transfer the data to the abnormal / alarm zone to achieve safe shutdown protection against destructive and malicious faults and output a shutdown command to the outside.

[0098] This embodiment employs a threshold dynamic escape mechanism and a dual-time-window fast-slow cascaded dejittering method. When a transient pulse arrives, the threshold dynamic escape mechanism instantaneously raises the threshold, thus increasing the sampled value of the pulse electrical signal. It did not exceed the dynamic threshold after the elevation. Therefore, in the state statistics of dual-time-window fast and slow cascaded de-jitter, noise is not included in the out-of-range samples, and the proportion of out-of-range samples in the fast window is... It remains at 0, and no erroneous state transition occurs. This achieves the combined and coordinated use of "physical constraint filtering of large pulses" and "state machine window filtering of small fluctuations".

[0099] While using the threshold dynamic escape mechanism alone can filter out steep single-point clutter, when there is high-frequency small-amplitude random noise or tiny fluctuations at the threshold edge, the system will generate alarm oscillations due to frequent touches of the judgment boundary because there is no statistical debouncing of the time-domain window.

[0100] Using dual-time-window fast-slow cascaded debouncing alone can be problematic when encountering short-duration, extremely large transient impulse noise (such as surges generated by the opening and closing of high-power electrical equipment or impacts from hard objects around the equipment). Although the impulse is not a mechanical fault, its extremely high amplitude within the short-time window will cause almost all sampling points to exceed the reference threshold. This will lead to The system instantly approaches 100%, bypassing the long-term slow confirmation window, which falsely triggers the emergency stop alarm, thus generating a false alarm.

[0101] The embodiments of the present invention can balance early rapid response and long-term wear stability, while saving memory overhead.

[0102] In summary, by extracting vibration energy under known steady-state operating conditions and using a closed-loop correction method to adjust the voltage alarm threshold, the risk of missed alarms due to hardware aging is reduced. A dynamic exponential escape mechanism for the threshold is constructed using first- and second-order differential change rates in the time domain and physical constraints, which can suppress the probability of false alarms caused by transient noise. At the lower-level microprocessor, a dual-time-window fast-slow cascade debouncing method is adopted, which can balance long-term degradation stability with immediate response to severe and malignant faults, avoiding noise-induced false alarms, identifying genuine sudden early severe faults, and saving on the microprocessor's computational and memory overhead.

[0103] like Figure 3 As shown, based on the above-described adaptive monitoring method for a triaxial accelerometer, this embodiment of the invention discloses an adaptive monitoring device for a triaxial accelerometer, comprising:

[0104] The multi-channel signal acquisition module 600 is used to acquire the time-domain triaxial vibration voltage signal collected by the sensor;

[0105] The voltage sensitivity degradation estimation module 610 is used to calculate the total effective value electrical signal energy of the triaxial vibration voltage signal within a preset time window when the operating conditions are stable, and to calculate the voltage sensitivity degradation coefficient based on the total effective value electrical signal energy.

[0106] The reference threshold correction module 620 is used to perform closed-loop adaptive real-time adjustment of the preset physical vibration safety limit based on the voltage sensitivity degradation coefficient to obtain the reference voltage alarm threshold.

[0107] The transient clutter interference determination module 630 is used to calculate the first-order differential rate of change and the second-order differential rate of change based on the acceleration data of the sensor, and to determine whether there is external transient clutter interference based on the absolute value of the first-order differential rate of change, the absolute value of the second-order differential rate of change and the physical inertial response limit.

[0108] The dynamic threshold escape module 640 is used to calculate the initial escape rise when it is determined to be external transient clutter interference, and to construct an escape attenuation model for calculating the dynamic voltage alarm threshold based on the initial escape rise.

[0109] The cascaded de-jitter state machine module 650 is used to count the proportion of samples in which the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold within a short-time fast window and a long-time slow window, and to set the current state interval based on the sample proportion.

[0110] In some implementations, the cascaded de-jitter state machine module includes a percentage calculation unit and a determination unit. The percentage calculation unit is used to calculate the percentage of samples in a short-term fast window where the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold to obtain a first percentage, and to calculate the percentage of samples in a long-term slow window where the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold to obtain a second percentage. The determination unit is used to set the current state interval as a region of interest when the first percentage > a first threshold; when the state interval is a region of interest, if the second percentage > a second threshold and the duration exceeds a preset duration, or if the first percentage > a third threshold, the current state interval is set as an abnormal region.

[0111] like Figure 4 As shown in the present invention, an electronic device is disclosed, including a memory 401 storing executable program code and a processor 402 coupled to the memory 401;

[0112] The processor 402 calls the executable program code stored in the memory 401 to execute the triaxial accelerometer adaptive monitoring method described in the above embodiments.

[0113] This invention also discloses a computer-readable storage medium storing a computer program that causes a computer to execute the triaxial accelerometer adaptive monitoring method described in the above embodiments.

[0114] The purpose of the above embodiments is to reproduce and derive the technical solution of the present invention by way of example, and to fully describe the technical solution, purpose and effect of the present invention. The purpose is to enable the public to have a more thorough and comprehensive understanding of the disclosure of the present invention, and not to limit the scope of protection of the present invention.

[0115] The above embodiments are not an exhaustive list based on the present invention, and there may be many other embodiments not listed. Any substitutions and improvements made without departing from the concept of the present invention are within the protection scope of the present invention.

Claims

1. An adaptive monitoring method for a triaxial accelerometer, characterized in that, include: Acquire the time-domain triaxial vibration voltage signal collected by the sensor; When operating under stable conditions, calculate the total effective value of the triaxial vibration voltage signal energy within a preset time window, and calculate the voltage sensitivity degradation coefficient based on the total effective value of the voltage signal energy. The preset physical vibration safety limit is adjusted in real time and closed-loop adaptively based on the voltage sensitivity degradation coefficient to obtain the reference voltage alarm threshold. The first-order differential rate of change and the second-order differential rate of change are calculated based on the acceleration data from the sensor. Based on the absolute values ​​of the first-order differential rate of change, the second-order differential rate of change, and the physical inertial response limit, it is determined whether the current situation is due to external transient clutter interference. When the external transient clutter interference is determined, the initial escape rise is calculated, and an escape attenuation model for calculating the dynamic voltage alarm threshold is constructed based on the initial escape rise. The proportion of samples in which the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold within the short-time fast window and the long-time slow window is statistically analyzed, and the current state interval is set according to the sample proportion.

2. The adaptive monitoring method as described in claim 1, characterized in that, The proportion of vibration voltage signals exceeding the dynamic voltage alarm threshold within the short-time fast window and long-time slow window is statistically analyzed. Based on this proportion, the current state interval is defined, including: The proportion of samples in which the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold within a short-time window is statistically analyzed to obtain the first proportion; The proportion of samples in which the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold within a long-term slow window is statistically analyzed to obtain the second proportion; When the first percentage is greater than the first threshold, the current state interval is set as the attention zone; when the state interval is the attention zone, when the second percentage is greater than the second threshold and the duration exceeds the preset duration, or when the first percentage is greater than the third threshold, the current state interval is set as the abnormal zone.

3. The adaptive monitoring method as described in claim 2, characterized in that, The target percentage is obtained by statistically analyzing the proportion of triaxial vibration voltage signals exceeding the dynamic voltage alarm threshold within the target window. The target window is either a short-time fast window or a long-time slow window, and the target percentage is either a first percentage or a second percentage, including: Based on whether the triaxial vibration voltage signal at each sampling moment within the target window exceeds the dynamic voltage alarm threshold at that sampling moment, a flag corresponding to each sampling moment is set to obtain a flag queue; The target percentage is obtained by performing a moving average on the flag queue.

4. The adaptive monitoring method as described in claim 1, characterized in that, The expression for calculating the total effective value of the triaxial vibration voltage signal energy within a preset time window is as follows: in, The total effective value of the electrical signal energy. This represents the total number of voltage signals within a preset time window. , , For the first A triaxial vibration voltage signal; The expression for calculating the voltage sensitivity degradation coefficient based on the total effective value of the electrical signal energy is as follows: in, for Voltage sensitivity degradation coefficient at time t. The smoothing time constant has a value ranging from 0.95 to 0.

999. This refers to the sensor's factory-specified sensitivity. These are the reference physical constants for steady-state vibration.

5. The adaptive monitoring method as described in claim 1, characterized in that, A triaxial accelerometer is mounted on the surface of the motor housing to determine whether the motor is operating under stable conditions, including: The motor is monitored in real time to obtain the average active power and average speed. The active power fluctuation rate is calculated based on the average active power and the speed fluctuation rate is calculated based on the average speed. When the active power fluctuation rate is lower than the first prior threshold and the speed fluctuation rate is lower than the second prior threshold, it is determined to be in a stable operating condition.

6. The adaptive monitoring method as described in claim 1, characterized in that, The expression for the escape decay model is: in, This is the initial escape lift. The reference voltage alarm threshold, To determine the moment when external transient clutter interference is triggered, The escape decay time constant has a value range of 1ms to 1000ms.

7. A triaxial accelerometer adaptive monitoring device, characterized in that, include: A multi-channel signal acquisition module is used to acquire time-domain triaxial vibration voltage signals collected by the sensor; The voltage sensitivity degradation estimation module is used to calculate the total effective value electrical signal energy of the triaxial vibration voltage signal within a preset time window when the operating conditions are stable, and to calculate the voltage sensitivity degradation coefficient based on the total effective value electrical signal energy. The reference threshold correction module is used to perform closed-loop adaptive real-time adjustment of the preset physical vibration safety limit based on the voltage sensitivity degradation coefficient to obtain the reference voltage alarm threshold. The transient clutter interference determination module is used to calculate the first-order differential rate of change and the second-order differential rate of change based on the acceleration data of the sensor, and to determine whether there is external transient clutter interference based on the absolute value of the first-order differential rate of change, the absolute value of the second-order differential rate of change and the physical inertial response limit. The dynamic threshold escape module is used to calculate the initial escape rise when it is determined to be external transient clutter interference, and to construct an escape attenuation model for calculating the dynamic voltage alarm threshold based on the initial escape rise. The cascaded de-jitter state machine module is used to count the proportion of samples in the short-time fast window and the long-time slow window where the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold, and to set the current state interval based on the sample proportion.

8. The adaptive monitoring device as described in claim 7, characterized in that, The cascaded anti-jitter state machine module includes a percentage calculation unit and a judgment unit. The percentage calculation unit is used to calculate the percentage of samples in a short-term fast window where the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold to obtain a first percentage, and to calculate the percentage of samples in a long-term slow window where the triaxial vibration voltage signal exceeds the dynamic voltage alarm threshold to obtain a second percentage. The judgment unit is used to set the current state interval as the attention zone when the first percentage > the first threshold; when the state interval is the attention zone, if the second percentage > the second threshold and the duration exceeds a preset duration, or if the first percentage > the third threshold, the current state interval is set as the abnormal zone.

9. An electronic device, characterized in that, It includes a memory storing executable program code and a processor coupled to the memory; the processor calls the executable program code stored in the memory to execute the triaxial accelerometer adaptive monitoring method according to any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein the computer program causes a computer to perform the triaxial accelerometer adaptive monitoring method according to any one of claims 1 to 6.

Images