A method and system for detecting pull-in voltage based on time-series data characteristics

Abstract

This invention belongs to the field of pull-in voltage testing technology, specifically relating to a pull-in voltage detection method and system based on time-series data characteristics. The method includes: acquiring the voltage and current time-series sequences of an electromagnetic drive component within a single energizing excitation cycle; obtaining the electromagnetic transient response factor at each sampling moment based on the characteristic differences between adjacent sampling moments in the voltage and current time-series sequences; obtaining the action characteristic confidence index at each sampling moment based on the electromagnetic transient response factor within a sliding window, combined with the current time-series sequence; determining the target action moment and target response voltage based on the action characteristic confidence index at each sampling moment; and obtaining the hysteresis index based on the values ​​of the stable maintenance phase and the aforementioned characteristics. This invention achieves high-precision equipment health status determination under complex power supply and high-temperature environment multi-source interference.

Description

Technical Field

[0001] This invention relates to the field of pull-in voltage testing technology. More specifically, this invention relates to a pull-in voltage detection method and system based on time-series data characteristics. Background Technology

[0002] As a core actuator in the starting systems of commercial heavy-duty trucks and industrial power equipment, the engine starter plays a vital role in fields such as combined highway transportation and harsh mining operations. During long-term, high-frequency starting or exposure to extreme high and low temperature alternating environments, the engine starter solenoid switch often experiences hidden problems such as internal coil insulation aging, moving and stationary contact erosion, and assembly interference of the moving iron core. Therefore, accurate detection of the solenoid switch's engagement voltage is crucial for ensuring the efficient and reliable operation of the entire vehicle's power system.

[0003] In related technologies, for example, Chinese patent application document with publication number CN106033111A discloses a method and system for diagnosing starter relay jamming. In the case of two starter relays connected in series with the starter, the method determines that the disconnected starter relay has jammed by engaging the other starter relay when one starter relay is disconnected and reading the voltage applied across the starter. If the voltage across the starter is not zero, the method determines that the disconnected starter relay has jammed.

[0004] However, in actual industrial power supply and operation environments, external power systems often experience voltage fluctuations due to sudden load changes, while the equipment control room usually experiences certain ambient temperature variations. Related technologies rely solely on reading absolute electrical values ​​to determine sluggishness, and cannot distinguish whether the abnormal electrical characteristics collected are physical manifestations caused by internal mechanical movements or noise interference caused by severe fluctuations in external power. Summary of the Invention

[0005] To address the technical problems of false alarms and low accuracy in mechanical jamming detection that easily occur in electromagnetic drive components under complex power supply networks and high-temperature environments, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a pull-in voltage detection method based on time-series data features, comprising: acquiring a voltage time-series sequence and a current time-series sequence of an electromagnetic drive component within a single energizing excitation cycle; acquiring an electromagnetic transient response factor at each sampling moment based on the characteristic differences between adjacent sampling moments in the voltage and current time-series sequences; extracting the fluctuation characteristics of the electromagnetic transient response factor in the local neighborhood corresponding to each sampling moment based on a sliding window, and combining the global features in the current time-series sequence to acquire an action feature confidence index at each sampling moment; taking the moment corresponding to the extreme value in the action feature confidence index as the target action moment, and extracting the corresponding target response voltage in the voltage time-series sequence based on the target action moment; acquiring the real-time thermal internal resistance based on the steady-state values ​​of the voltage and current time-series sequences during the stable maintenance phase, acquiring a hysteresis index by combining the target action moment, the target response voltage, and reference parameters, and performing a mechanical hysteresis warning for the electromagnetic drive component based on the hysteresis index.

[0007] This invention utilizes the difference in voltage variation between external power supply fluctuations and internal back electromotive force, as well as the difference in temporal distribution between continuous displacement characteristics caused by nonlinear distribution of friction and local extrema points caused by high-frequency white noise, to extract masked weak back electromotive force characteristics. Simultaneously, it uses real-time thermal internal resistance obtained at the end of power-on to interrupt the judgment path caused by simple time delays due to high-temperature environments, reducing interference from ripple in industrial DC power supply systems and voltage fluctuations caused by load mutations. This reduces the possibility of misjudging normal action delays as mechanical faults in the control cabin of industrial equipment at extremely high ambient temperatures, thereby improving the accuracy of detecting internal mechanical jamming states of electromagnetic drive components under complex power supply environments and high-temperature conditions.

[0008] Preferably, before obtaining the electromagnetic transient response factor of each sampling moment based on the characteristic differences between adjacent sampling moments in the voltage time series and current time series, the method includes: controlling the electromagnetic drive component to be in a stable operating condition at room temperature during the factory calibration or power-on self-test stage of the electromagnetic drive component; applying a standard rated drive signal to the electromagnetic drive component; and recording the average readings of the current sensor and voltage sensor during the stable maintenance stage to obtain the reference rated current and reference rated voltage, respectively.

[0009] Preferably, the electromagnetic transient response factor satisfies the following relationship: In the formula, For the first Electromagnetic transient response factor at each sampling time, The first in the current time sequence The sampling time and the first The absolute value of the difference between the current values ​​at each sampling time. The voltage time sequence is the first... The sampling time and the first The absolute value of the difference between the voltage values ​​at each sampling time. The reference rated current is pre-calibrated under normal temperature conditions. This is the reference rated voltage pre-calibrated under normal temperature conditions.

[0010] This invention constructs an exponential suppression weight by calculating the relative rate of change of voltage and the voltage fluctuation penalty coefficient, which distinguishes whether the local current dip is caused by internal mechanical motion or external power supply interference. Thus, it extracts the true electromagnetic transient performance under the interference of external power supply fluctuations, and improves the anti-interference ability of subsequent mechanical jamming feature extraction.

[0011] Preferably, the step of extracting the fluctuation characteristics of the electromagnetic transient response factor in the local neighborhood corresponding to each sampling time based on the sliding window, and combining the global features in the current time series to obtain the action feature confidence index for each sampling time includes: taking the current sampling time as the center, extracting each sampling time within a sliding window of a preset length as a neighborhood time series; calculating the standard deviation of the electromagnetic transient response factor of all sampling times in the neighborhood time series; traversing the current time series to find the maximum current value; and combining the standard deviation, the maximum current value, the electromagnetic transient response factor at the current sampling time, and the current value at the current sampling time to obtain the action feature confidence index for the current sampling time.

[0012] This invention extracts a sliding window centered on the current sampling time as a neighborhood time sequence, calculates the standard deviation of the electromagnetic transient response factor for all sampling times in the neighborhood time sequence, and obtains the action feature confidence index by combining the maximum current value of the current time sequence. This transforms isolated sampling point values ​​into a local fluctuation distribution state within a time interval, which aligns with the nonlinear distribution of internal friction force in the electromagnetic drive component during actual displacement, resulting in current dip characteristics distributed within a short time interval. This reduces the possibility of local extreme points being misjudged as action segments due to high-frequency random white noise, thus improving the accuracy of action feature extraction.

[0013] Preferably, the action feature confidence index satisfies the following relationship: In the formula, For the first Confidence index of action features at each sampling time For the first Electromagnetic transient response factor at each sampling time, For the first The standard deviation of all electromagnetic transient response factors in the neighborhood time sequence corresponding to each sampling time. The first in the current time sequence The current value at each sampling time. This represents the maximum current value in the current time series. A preset constant is used to prevent the denominator from being zero.

[0014] This invention enables the acquisition of a larger confidence index for action characteristics when the local fluctuation is significant and the current is below the maximum current value, and a smaller confidence index for action characteristics when the local fluctuation is less significant or in the current saturation steady state at the end of the power-on period. This utilizes the global relative current depth at the beginning of power-on to constrain the local fluctuation performance, making the moment when the internal components perform physical actions stand out in the continuous time-series data stream, providing reliable data support for determining the moment when mechanical displacement occurs.

[0015] Preferably, obtaining the real-time thermal internal resistance based on the steady-state values ​​of the voltage and current time series during the stable maintenance phase includes: extracting all voltage and current sampling point values ​​falling within the stable maintenance phase from the voltage and current time series, respectively. The stable maintenance phase is the period from after the electromagnetic drive component completes its mechanical transient displacement until the power-on excitation command ends; calculating the average value of the extracted voltage sampling point values ​​and using the average value of the extracted voltage sampling point values ​​as the steady-state voltage data; calculating the average value of the extracted current sampling point values ​​and using the average value of the extracted current sampling point values ​​as the steady-state current data; and obtaining the real-time thermal internal resistance based on the steady-state voltage and steady-state current data.

[0016] Preferably, the step of obtaining the hysteresis index by combining the target action time, target response voltage, and reference parameters includes: calculating the absolute value of the difference between the target response voltage and the reference response voltage, and using the ratio of the absolute value of the difference to the reference response voltage as the voltage deviation rate; calculating the absolute value of the difference between the target action time and the room temperature reference action time, and using the ratio of the absolute value of the difference to the room temperature reference action time as the action time delay rate; calculating the absolute value of the difference between the real-time hot internal resistance and the room temperature reference internal resistance, and using the ratio of the absolute value of the difference to the room temperature reference internal resistance as the internal resistance drift rate; and obtaining the hysteresis index by combining the voltage deviation rate, the action time delay rate, and the internal resistance drift rate.

[0017] This invention extracts the action feature confidence index with the largest value from the sequence data, marks the sampling time corresponding to the action feature confidence index with the largest value as the target action time, and retrieves the voltage value at the coordinates corresponding to the target action time from the voltage time series as the target response voltage. Based on the removal of interference features caused by external power supply noise and the preservation of real mechanical motion features by the action feature confidence index, the moment when the internal components undergo mechanical displacement is determined by locating the maximum value. This realizes the operation of extracting the external voltage at the moment of mechanical displacement from the original voltage sequence containing random noise, and improves the accuracy of extracting the target response voltage.

[0018] Preferably, the acquisition of the voltage timing sequence and current timing sequence of the electromagnetic drive component within a single power-on excitation cycle includes: during the power-on initialization phase of the electromagnetic drive component, controlling the electromagnetic drive component to perform a single complete power-on and power-off action; within the single power-on excitation cycle, synchronously acquiring the voltage signal at both ends of the electromagnetic drive component and the current signal flowing through the coil according to a preset sampling frequency; and constructing a voltage timing sequence and a current timing sequence containing multiple consecutive sampling points arranged in chronological order.

[0019] Preferably, the step of taking the time corresponding to the extreme value in the action feature confidence index as the target action time and extracting the corresponding target response voltage in the voltage time series based on the target action time includes: acquiring sequence data containing the action feature confidence index of all sampling times; extracting the action feature confidence index with the largest value in the sequence data, taking the sampling time corresponding to the action feature confidence index with the largest value as the target action time; extracting the voltage value at the coordinate corresponding to the target action time in the voltage time series, and taking the extracted voltage value as the target response voltage.

[0020] Secondly, the present invention provides a pull-in voltage detection system based on timing data characteristics, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned pull-in voltage detection method based on timing data characteristics is implemented.

[0021] By adopting the above technical solution, a computer program for detecting pull-in voltage based on time-series data characteristics is generated and stored in a memory for loading and execution by a processor. This allows for the creation of a terminal device based on the memory and processor, facilitating its use.

[0022] The beneficial effects of this invention are as follows: This invention numerically corrects the abrupt change characteristics of the current time sequence by synchronously acquiring the voltage time sequence, and uses the steady-state electrical performance at the end of the energization period to compensate for the time delay of the transient displacement process, distinguishing between external power supply fluctuation interference and the actual internal mechanical displacement performance. At the same time, it separates the increase in coil internal resistance caused by changes in ambient temperature from the mechanical friction resistance caused by internal wear, reducing misjudgments of the health status of electromagnetic drive components in complex industrial power supply networks and high-temperature environments, and improving the accuracy of detecting the mechanical jamming state of electromagnetic drive components.

[0023] This invention extracts the electromagnetic transient response and motion characteristic fluctuation of electromagnetic drive components within the operating cycle through continuous time-series data. Combined with real-time thermal internal resistance to obtain the hysteresis index, it transforms instantaneous mechanical displacement into data changes that include complete electrical state changes and local fluctuation distribution characteristics. This reduces the interference of high-frequency electrical noise and load abrupt changes on the feature extraction process, ensuring that the extracted data features truly reflect the process of electromagnetic drive components overcoming internal static resistance, thereby improving the reliability of equipment condition assessment results.

[0024] This invention integrates the electromagnetic force establishment state of the electromagnetic drive component at the moment of mechanical transient displacement and the heating state at the end of the energization period, eliminating false abnormal features caused by external voltage fluctuations and high temperature baking. It improves the anti-interference capability of monitoring the status of the underlying drive components during the operation of industrial equipment, and helps relevant personnel to promptly detect internal wear or spring aging of the electromagnetic drive component in complex industrial operating environments. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating a pull-in voltage detection method based on time-series data features according to the present invention; Figure 2 This is a schematic diagram illustrating the changes in the confidence index of action characteristics. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0028] This invention discloses a pull-in voltage detection method based on time-series data features, referring to... Figure 1 This includes steps S1-S5: S1. Obtain the voltage timing sequence and the current timing sequence.

[0029] It should be noted that the critical response voltage at the moment of mechanical transient displacement of the electromagnetic drive component is an important parameter for evaluating its internal mechanical jamming and spring fatigue aging. Due to the strong ripple and voltage fluctuation interference caused by load changes in industrial DC power supply systems, the voltage across the electromagnetic drive component and the current flowing through the coil directly collected contain a large amount of random noise, which masks the weak back electromotive force characteristics reflecting the internal mechanical motion. Therefore, synchronously acquiring complete voltage and current time-series data through high-frequency sampling provides underlying data support for subsequent disturbance rejection characteristic mapping.

[0030] Specifically, during the power-on initialization phase, the electromagnetic drive component is controlled to perform a single complete power-on and power-off action. Within this single power-on excitation cycle, the voltage signal at both ends of the electromagnetic drive component and the current signal flowing through the coil are synchronously acquired according to a preset sampling frequency, and voltage timing sequence and current timing sequence containing multiple consecutive sampling points arranged in chronological order are constructed respectively.

[0031] For example, the preset sampling frequency is 50kHz.

[0032] S2. Obtain the electromagnetic transient response factor at each sampling time.

[0033] It should be noted that when the internal mechanism of the electromagnetic drive component undergoes transient mechanical displacement, the internal air gap decreases sharply, causing a sudden change in coil inductance. This results in a localized dip in the back electromotive force on the current time-series curve. Due to the limitations of harsh industrial power supply networks, drastic voltage fluctuations from the external power source can also cause similar current dips. Conventional single-variable current differentiation methods cannot distinguish whether this change originates from internal mechanical motion or external power interference. Therefore, this invention uses the voltage change rate at adjacent moments as a constraint and dynamically adjusts the confidence weight of the current change rate, effectively suppressing spurious current abrupt changes caused by external power fluctuations.

[0034] Specifically, for the first current time series... At the sampling time, calculate its relationship with the first sampling time. The absolute value of the current difference at the nth sampling time, for the nth sampling time sequence. At the sampling time, calculate its relationship with the first sampling time. The absolute value of the voltage difference at each sampling time. Retrieve the pre-calibrated reference rated current and reference rated voltage, and combine the absolute value of the current difference, the reference rated current, the absolute value of the voltage difference, and the reference rated voltage to obtain the [missing value]. Electromagnetic transient response factor at each sampling time.

[0035] For example, the method for obtaining the reference rated current and reference rated voltage includes: during the factory calibration or power-on self-test phase of the electromagnetic drive component, when the control component is in a stable operating condition at room temperature, a standard rated drive signal is applied, and the average readings of the current sensor and voltage sensor during the stable maintenance phase are recorded to obtain the reference rated current and reference rated voltage.

[0036] Specifically, the electromagnetic transient response factor satisfies the following relationship: ; In the formula, For the first Electromagnetic transient response factor at each sampling time, The first in the current time sequence The sampling time and the first The absolute value of the difference between the current values ​​at each sampling time. The voltage time sequence is the first... The sampling time and the first The absolute value of the difference between the voltage values ​​at each sampling time. The reference rated current is pre-calibrated under normal temperature conditions. This is the reference rated voltage pre-calibrated under normal temperature conditions.

[0037] in, This represents the relative rate of change of current between adjacent sampling times. The larger the value, the greater the probability of a sudden change in current at that moment, resulting in a larger base value for the electromagnetic transient response factor. The smaller the value, the smoother the current change at that moment, resulting in a smaller base value for the electromagnetic transient response factor. Represents the exponential suppression weights resulting from voltage fluctuations, and the rate of change of voltage in this voltage time series. The larger the value, the more likely the current fluctuation at that moment was caused by external power source interference, leading to... The closer it gets to 0, the more it affects... The greater the attenuation, the more effectively the spurious current abrupt characteristics are suppressed. The smaller the value, the more likely the current fluctuation at that moment originates from the real back electromotive force generated by internal mechanical displacement, leading to... The closer it is to 1, the more the pure electromagnetic transient characteristics are preserved in the electromagnetic transient response factor, thus through... right The electromagnetic transient response factor is obtained by performing downward correction.

[0038] S3. Obtain the confidence index of action features at each sampling time.

[0039] It should be noted that the internal friction force of the electromagnetic drive component exhibits a nonlinear distribution during actual displacement, causing the current dip feature to be distributed over a very short time interval rather than a single isolated sampling point. To avoid misjudging local singularities caused by high-frequency random white noise, this invention employs a sliding time window to extract local saliency and uses the global relative current depth at the initial stage of energization as a constraint, thus highlighting the actual physical action segment in the entire time-series data stream.

[0040] Specifically, with the first Centered on a sampling time point, each sampling time point within a sliding window of a preset length is selected as a neighborhood time sequence. The standard deviation of the electromagnetic transient response factor for all sampling times in this neighborhood time sequence is calculated. The current time series is traversed to find the maximum current value, and the maximum current value is determined by combining the standard deviation of the electromagnetic transient response factor with the sampling time point. The electromagnetic transient response factor, maximum current value, and the first sampling time are all present. The current value at the sampling time is obtained. Confidence index of action features at each sampling time.

[0041] For example, the preset length is 51 sampling points.

[0042] Specifically, the confidence index of action features satisfies the following relationship: ; In the formula, For the first Confidence index of action features at each sampling time For the first Electromagnetic transient response factor at each sampling time, For the first The standard deviation of all electromagnetic transient response factors in the neighborhood time sequence corresponding to each sampling time. The first in the current time sequence The current value at each sampling time. This represents the maximum current value in the current time series. To prevent the default constant where the denominator is zero, this embodiment... .

[0043] in, This represents the significance of the current sampling time relative to the fluctuations within the local window. This represents the relative depth of the current value compared to the global maximum current value, indicating the significance of local fluctuations. The larger the value and the more significantly the current is compared to the maximum current value, the greater the likelihood of displacement of the moving parts inside the electromagnetic drive assembly at that moment, resulting in a higher confidence index for the acquired motion characteristics; the greater the significance of local fluctuations. The smaller the current value, or the more likely the current is to be in the final stage of steady-state saturation, the lower the probability of displacement. This ultimately leads to the confidence index of the action characteristic approaching 0. Therefore, by... right Develop confidence indices for action features by performing synergistic interactions.

[0044] For example, Figure 2 This diagram illustrates the change in the confidence index of the action feature in this invention. It shows the normalized amplitude of the derivative of the conventional current and the confidence index of the action feature. As can be seen, when faced with false current dips caused by strong external power fluctuations, the derivative curve of the conventional current exhibits large, violent oscillations and prominent false peaks, easily leading to misjudgments. In contrast, the confidence index of the action feature extracted by this invention shows its value attenuated downwards in this interference range. Conversely, at the moment when the electromagnetic drive component experiences a real mechanical transient displacement, the current surge caused by the back electromotive force is preserved and nonlinearly amplified, causing the confidence index of the action feature to rise rapidly. This effectively separates the real action feature from high-frequency noise and interference background, ultimately anchoring the target action moment and improving the robustness of feature extraction in harsh industrial environments.

[0045] S4. Determine the target action time and obtain the target response voltage.

[0046] It should be noted that the motion feature confidence index enhances the realistic mechanical motion characteristics and effectively suppresses interference characteristics caused by external power supply noise. By directly locating the global extreme point of the motion feature confidence index within the entire time-series cycle, the absolute time coordinate of the mechanical displacement of internal components can be accurately anchored, thereby extracting the corresponding external voltage from the original voltage sequence.

[0047] Specifically, acquire sequence data containing the confidence indices of action features at all sampling times. Extract the largest confidence index of action features from the sequence data and mark the sampling time corresponding to this confidence index as the target action time. Retrieve the voltage time series and extract the voltage value at the coordinates corresponding to the target action time, using this voltage value as the target response voltage.

[0048] S5. Obtain the lag index.

[0049] It should be noted that the extremely high ambient temperature inside the control cabin of industrial equipment causes a significant increase in the copper wire resistance of the electromagnetic coil. Typically, this increased thermal resistance slows down the establishment of electromagnetic force, resulting in a substantial delay in the action time and a drift in the target response voltage, even if the internal mechanical structure of the component is healthy. To prevent misjudging normal time delays caused solely by high temperatures as mechanical jamming faults, this invention obtains a jamming index based on the values ​​of the stable maintenance phases of the voltage and current timing sequences, as well as the target action time and target response voltage.

[0050] Specifically, steady-state voltage and current data at the end of the energizing period are acquired from the voltage and current time series. Based on this steady-state voltage and current data, the real-time hot-state internal resistance is obtained. Pre-calibrated reference response voltage, room-temperature reference action time, and room-temperature reference internal resistance are acquired. The absolute value of the difference between the target response voltage and the reference response voltage is calculated, and the ratio of this absolute value to the reference response voltage is the voltage deviation rate. The absolute value of the difference between the target action time and the room-temperature reference action time is calculated, and the ratio of this absolute value to the room-temperature reference action time is the action time delay rate. The absolute value of the difference between the real-time hot-state internal resistance and the room-temperature reference internal resistance is calculated, and the ratio of this absolute value to the room-temperature reference internal resistance is the internal resistance drift rate. The hysteresis index is obtained by combining the voltage deviation rate, action time delay rate, and internal resistance drift rate.

[0051] For example, the final stage of power-on is the stable maintenance phase after the electromagnetic drive component completes its mechanical transient displacement until the power-on excitation command ends. The duration of this phase is typically 10% to 20% of the entire power-on cycle. The steady-state voltage and current data at the final stage of power-on are obtained by: extracting all voltage and current sampling point values ​​that fall within the stable maintenance phase from the voltage time series and current time series, respectively; calculating the average value of the extracted voltage sampling point values, and using this average value as the steady-state voltage data; and calculating the average value of the extracted current sampling point values, and using this average value as the steady-state current data.

[0052] For example, the methods for obtaining the reference response voltage, the reference action time at room temperature, and the reference internal resistance at room temperature include: performing multiple power-on tests on a group of fault-free sample components under a constant temperature environment of standard atmospheric pressure and 25 degrees Celsius, recording the voltage, time, and steady-state resistance at the moment of displacement, and calculating the corresponding average values ​​to obtain the reference response voltage, the reference action time at room temperature, and the reference internal resistance at room temperature.

[0053] Specifically, the hysteresis index satisfies the following relationship: ; In the formula, The lag index, Voltage deviation rate, For action time delay rate, This represents the internal resistance drift rate.

[0054] in, The larger the value, the greater the likelihood that the critical point at which the electromagnetic force overcomes the internal static resistance will shift. The larger the value, the greater the likelihood that the armature's movement will be obstructed. Utilizing the synergistic effect of objective phenomena, when abnormal voltage drift and severe lag occur simultaneously, This can lead to a sharp increase in the lag index, thereby amplifying the actual lag and deterioration characteristics. This represents the thermal compensation factor used to counteract temperature rise interference, based on the principle that coil resistance increases with temperature, when the internal resistance drift rate... When the value increases, it indicates that the component is operating under extreme high-temperature conditions. At this point, the thermal compensation factor approaches 0, which has a significant impact on... The greater the reduction, the better to avoid the artificially inflated jamming index caused by high-temperature delay; when the internal resistance drift rate The smaller the value, the more likely the component is operating at room temperature, and the thermal compensation factor is close to 1, making it more efficient. The hysteresis index was fully preserved, enabling the determination of mechanical health under different ambient temperatures.

[0055] Furthermore, in response to the jamming index exceeding a preset alarm tolerance threshold, a mechanical jamming warning signal for the drive component is generated, thereby achieving high-precision equipment health status determination under complex multi-source interference, for example, an alarm tolerance threshold of 0.65.

[0056] This invention also discloses a pull-in voltage detection system based on timing data features, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a pull-in voltage detection method based on timing data features according to the present invention.

[0057] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. A method for detecting pull-in voltage based on time-series data characteristics, characterized in that, include: Obtain the voltage timing sequence and current timing sequence of the electromagnetic drive component during a single energizing cycle; The electromagnetic transient response factor at each sampling moment is obtained based on the characteristic differences between adjacent sampling moments in the voltage time series and the current time series. Based on the sliding window, the fluctuation characteristics of the electromagnetic transient response factor in the local neighborhood corresponding to each sampling time are extracted, and combined with the global features in the current time series, the confidence index of the action feature at each sampling time is obtained. The time corresponding to the extreme value in the action feature confidence index is taken as the target action time, and the corresponding target response voltage is extracted from the voltage time series based on the target action time; The real-time thermal internal resistance is obtained by taking steady-state values ​​of the voltage and current time series during the stable maintenance phase. The hysteresis index is obtained by combining the target action time, target response voltage and reference parameters. The mechanical hysteresis warning of the electromagnetic drive component is then based on the hysteresis index.

2. The pull-in voltage detection method based on time-series data features according to claim 1, characterized in that, Before obtaining the electromagnetic transient response factor for each sampling moment based on the characteristic differences between adjacent sampling moments in the voltage time series and current time series, the process includes: During the factory calibration or power-on self-test phase of the electromagnetic drive assembly, the electromagnetic drive assembly is controlled to operate under stable conditions at room temperature; a standard rated drive signal is applied to the electromagnetic drive assembly; the average readings of the current sensor and voltage sensor during the stable maintenance phase are recorded to obtain the reference rated current and reference rated voltage, respectively.

3. The pull-in voltage detection method based on time-series data features according to claim 1 or 2, characterized in that, The electromagnetic transient response factor satisfies the following relationship: ； In the formula, For the first Electromagnetic transient response factor at each sampling time, The first in the current time sequence The sampling time and the first The absolute value of the difference between the current values ​​at each sampling time. The voltage time sequence is the first... The sampling time and the first The absolute value of the difference between the voltage values ​​at each sampling time. The reference rated current is pre-calibrated under normal temperature conditions. This is the reference rated voltage pre-calibrated under normal temperature conditions.

4. The pull-in voltage detection method based on time-series data features according to claim 1, characterized in that, The method of extracting the fluctuation characteristics of the electromagnetic transient response factor in the local neighborhood corresponding to each sampling time based on the sliding window, and combining the global features in the current time series to obtain the action feature confidence index at each sampling time includes: taking the current sampling time as the center, extracting each sampling time within a sliding window of a preset length as the neighborhood time series; calculating the standard deviation of the electromagnetic transient response factor of all sampling times in the neighborhood time series; traversing the current time series to find the maximum current value; and combining the standard deviation, the maximum current value, the electromagnetic transient response factor at the current sampling time, and the current value at the current sampling time to obtain the action feature confidence index at the current sampling time.

5. The pull-in voltage detection method based on time-series data features according to claim 4, characterized in that, The confidence index of the action feature satisfies the following relationship: ； In the formula, For the first Confidence index of action features at each sampling time For the first Electromagnetic transient response factor at each sampling time, For the first The standard deviation of all electromagnetic transient response factors in the neighborhood time sequence corresponding to each sampling time. The first in the current time sequence The current value at each sampling time. This represents the maximum current value in the current time series. A preset constant is used to prevent the denominator from being zero.

6. The pull-in voltage detection method based on time-series data features according to claim 1, characterized in that, The method of obtaining real-time thermal internal resistance based on steady-state values ​​during the stable maintenance phase of the voltage and current time series includes: extracting all voltage and current sampling point values ​​falling within the stable maintenance phase from the voltage and current time series, respectively. The stable maintenance phase is the period from after the electromagnetic drive component completes its mechanical transient displacement until the power-on excitation command ends; calculating the average value of the extracted voltage sampling point values ​​and using the average value of the extracted voltage sampling point values ​​as the steady-state voltage data; calculating the average value of the extracted current sampling point values ​​and using the average value of the extracted current sampling point values ​​as the steady-state current data; and obtaining the real-time thermal internal resistance based on the steady-state voltage and steady-state current data.

7. The pull-in voltage detection method based on time-series data features according to claim 1, characterized in that, The method of obtaining the hysteresis index by combining the target action time, target response voltage, and reference parameters includes: calculating the absolute value of the difference between the target response voltage and the reference response voltage, and using the ratio of the absolute value of the difference to the reference response voltage as the voltage deviation rate; calculating the absolute value of the difference between the target action time and the room temperature reference action time, and using the ratio of the absolute value of the difference to the room temperature reference action time as the action time delay rate; calculating the absolute value of the difference between the real-time hot internal resistance and the room temperature reference internal resistance, and using the ratio of the absolute value of the difference to the room temperature reference internal resistance as the internal resistance drift rate; and obtaining the hysteresis index by combining the voltage deviation rate, action time delay rate, and internal resistance drift rate.

8. The pull-in voltage detection method based on time-series data features according to claim 1, characterized in that, The step of taking the time corresponding to the extreme value in the action feature confidence index as the target action time and extracting the corresponding target response voltage in the voltage time series based on the target action time includes: acquiring sequence data containing the action feature confidence index of all sampling times; extracting the action feature confidence index with the largest value in the sequence data, taking the sampling time corresponding to the action feature confidence index with the largest value as the target action time; extracting the voltage value at the coordinate corresponding to the target action time in the voltage time series, and taking the extracted voltage value as the target response voltage.

9. The pull-in voltage detection method based on time-series data characteristics according to claim 1, characterized in that, The acquisition of the voltage timing sequence and current timing sequence of the electromagnetic drive component within a single energizing excitation cycle includes: During the power-on initialization phase of the electromagnetic drive component, the electromagnetic drive component is controlled to perform a single complete power-on and power-off action. Within a single power-on excitation cycle, the voltage signal at both ends of the electromagnetic drive component and the current signal flowing through the coil are synchronously acquired according to the preset sampling frequency. Voltage timing sequence and current timing sequence containing multiple consecutive sampling points arranged in chronological order are constructed respectively.

10. A pull-in voltage detection system based on time-series data characteristics, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a pull-in voltage detection method based on timing data characteristics according to any one of claims 1-9.

Images