Power equipment full-factor monitoring data analysis method fusing multiple rules

Abstract

The application discloses a power equipment full-element monitoring data analysis method fusing multiple rules, and relates to the technical field of data analysis, including the following steps: collecting current, temperature and vibration change sequences in the load sudden jump process of power equipment, constructing a phase change atlas and performing phase offset compensation to form the same phase change sequence; generating a dynamic priority adjustment table based on the phase consistency degree, rearranging the alarm triggering sequence; combining the abnormal duration interval to optimize the determination path and exclude the short transient recovery segment; and synchronously updating the triggering sequence according to the phase alignment mode when the load suddenly jumps again. The application constructs the phase change atlas of power equipment and performs phase offset compensation, realizes the alignment of the same phase change sequence, guarantees the consistency of the alarm triggering sequence and the load rhythm, avoids misjudging the abnormality as recovery, and simultaneously maintains the abnormal trend continuity through dynamic priority rearrangement and duration interval constraint, thereby reducing the risk accumulation hidden danger of power equipment.

Description

Technical Field

[0001] This application relates to the field of data analysis technology, specifically to a method for analyzing all elements of power equipment monitoring data that integrates multiple rules. Background Technology

[0002] Full-element monitoring data analysis of power equipment involves unified modeling and comprehensive analysis of multi-dimensional monitoring information generated during equipment operation. This encompasses key elements such as temperature, vibration, partial discharge, current, voltage, load, oil chromatography, infrared imaging, operational behavior records, and environmental parameters, systematically reflecting the equipment's structural state, performance evolution trajectory, and degradation trends. By constructing a unified time base, data from different sources and sampling frequencies are time-series aligned and correlated, accurately capturing state evolution paths, abnormal deviation characteristics, and potential fault modes, thus forming a comprehensive assessment of equipment health level, potential hazard location, and risk level. Multi-rule analysis, within the full-element data framework, introduces physical mechanism rules, empirical threshold rules, operational mode rules, and time-series correlation rules, enabling multiple judgment logics to work synergistically within the same data system. Through cross-comparison, chain constraints, and anomaly verification mechanisms, the ability to identify weak anomalies in complex load fluctuation and environmental disturbance scenarios can be enhanced, improving the accuracy of anomaly tracing and achieving stable and reliable diagnosis of the entire lifecycle operation status of power equipment.

[0003] The existing technology has the following shortcomings:

[0004] During the operation of power equipment, when the load undergoes a sudden change, the characteristic quantities upon which various judgment rules rely will experience phase shifts or even phase reversals within a very short time. If a stable dynamic synchronization relationship is not established between multiple rules, the alarm trigger chain originally used to identify anomalies may be misinterpreted as a state recovery path under the effect of phase reversal, causing dangerous signs to be filtered out as normal fluctuations. This phenomenon is more likely to occur in scenarios with rapid load switching and frequent power surges, easily leading to the continuous accumulation of high-risk states without timely identification, resulting in delayed protection actions, an expanded fault range, and even cascading damage to equipment.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this application is to provide a method for analyzing all-element monitoring data of power equipment that integrates multiple rules. By constructing a phase change spectrum of power equipment and performing phase offset compensation, the method achieves alignment of the same phase change sequence, ensuring that the alarm triggering order is consistent with the load rhythm and avoiding the misjudgment of anomalies as recovery. At the same time, by dynamically prioritizing and constraining continuous intervals, the method maintains the continuity of abnormal trends and reduces the potential risks of power equipment accumulation, thereby solving the problems mentioned in the background technology.

[0007] To achieve the above objectives, this application provides the following technical solution: a method for analyzing all elements of power equipment monitoring data that integrates multiple rules, comprising the following steps:

[0008] The current change sequence, temperature change sequence, and vibration change sequence of power equipment during load jump are collected. The current change sequence, temperature change sequence, and vibration change sequence are unfolded under a unified time coordinate. The phase start point of each change sequence is extracted, and a phase change spectrum is constructed based on each phase start point.

[0009] Based on the phase start point in the phase change spectrum, phase shift compensation processing is performed on the current change sequence, temperature change sequence and vibration change sequence respectively, and the phase start points are aligned to obtain the same phase change sequence.

[0010] Based on the same phase change sequence, the phase consistency between each change sequence is calculated. A dynamic priority adjustment table is generated based on the phase consistency. The original alarm triggering order is then reordered according to the dynamic priority adjustment table to form a triggering order corresponding to the load change rhythm.

[0011] Based on the dynamic priority adjustment table, the current decision path is rearranged in time, the continuous interval of abnormal features in the same phase change sequence is written into the trigger condition, and short instantaneous recovery segments are excluded from the decision path.

[0012] After the decision path is completed and the timing is rearranged, its evolution rhythm is continuously tracked based on the same phase change sequence. When the load changes abruptly again, the triggering order is updated synchronously according to the alignment method in the phase change spectrum to maintain the stability of the decision direction and prevent dangerous states from being identified as recovery states.

[0013] Preferably, constructing the phase change spectrum includes the following steps:

[0014] When the current value undergoes a step change between adjacent time scales and maintains the same direction of change within a continuous time scale, the load jump start time segment is determined, and the current change sequence, temperature change sequence, and vibration change sequence within the corresponding load jump start time segment are extracted.

[0015] Arrange the current change sequence, temperature change sequence, and vibration change sequence point by point on a unified time coordinate.

[0016] The numerical change trajectories of the current change sequence, temperature change sequence, and vibration change sequence are scanned respectively, the phase start point is recorded, and a phase record set containing three time scale values ​​is formed.

[0017] Based on the phase record set, the phase start points and corresponding change intervals of the current change sequence, temperature change sequence, and vibration change sequence are marked to form a phase change spectrum.

[0018] Preferably, obtaining the in-phase change sequence includes the following steps:

[0019] Read the time scale values ​​of the phase start points of the current change sequence, temperature change sequence and vibration change sequence in the phase change spectrum on the unified time coordinate, calculate the time interval between the time scale of each phase start point and the unified alignment reference time scale and record it as the time offset.

[0020] Based on the time offset, the current change sequence, temperature change sequence and vibration change sequence are all translated on the time scale under the unified time coordinate, and the phase starting point of each change sequence is moved to the unified alignment reference time scale position.

[0021] The current change sequence, temperature change sequence, and vibration change sequence that have completed the overall translation process are superimposed and arranged on a unified time coordinate. The change trajectories of each change sequence are synchronously unfolded with a unified alignment reference time scale as a common starting point, forming a change sequence in phase.

[0022] Preferably, the unified alignment reference time scale is selected as the earliest time scale among the phase start point time scales of the current change sequence, temperature change sequence, and vibration change sequence; the time offset is determined based on the time interval between each phase start point time scale and the unified alignment reference time scale; the same phase change sequence is judged by the coincidence of the phase start point under the unified time coordinate.

[0023] Preferably, generating a dynamic priority adjustment table based on the degree of phase consistency includes the following steps:

[0024] In the same-phase change sequence, a continuous time segment is selected starting from the unified aligned reference time scale. The change direction of the current change sequence, temperature change sequence and vibration change sequence is marked point by point and merged to form a continuous direction segment.

[0025] Based on the continuous directional segments, segment by segment comparison is performed under a unified time coordinate to record the phase-consistent segments and durations of current change sequences, temperature change sequences and vibration change sequences within the same time segment, forming a phase consistency record result.

[0026] Based on the results of the phase consistency record, the cumulative duration of each change sequence participating in the phase consistency segment is calculated and arranged in order of cumulative duration to generate a dynamic priority adjustment table.

[0027] Preferably, the phase consistency recording results are sorted based on the cumulative duration of the directional consistent segments of the current change sequence, temperature change sequence, and vibration change sequence under a unified time coordinate; the dynamic priority adjustment table records the priority order of each change sequence and the corresponding time segment position; the triggering order is arranged according to the priority order in the dynamic priority adjustment table.

[0028] Preferably, the temporal rearrangement of the current decision path based on the dynamic priority adjustment table includes the following steps:

[0029] Based on the dynamic priority adjustment table, the alarm conditions for current change sequence, temperature change sequence, and vibration change sequence in the original judgment path are reordered, and the reordered trigger conditions are matched with the abnormal sections of the same phase change sequence under the same time coordinate.

[0030] Mark the abnormal duration intervals of current change sequences, temperature change sequences, and vibration change sequences, and write the start time scale, end time scale, and duration of the abnormal duration interval into the corresponding trigger conditions.

[0031] Identify and record reverse change segments within the abnormal duration interval that have a duration less than a preset time scale as short instantaneous recovery segments, and exclude these short instantaneous recovery segments from the judgment process during the judgment path execution.

[0032] Preferably, the abnormal duration interval is determined based on the number of time scales that change continuously in one direction under a unified time coordinate, and the determination path is executed according to the triggering order in the dynamic priority adjustment table and in combination with the abnormal duration interval.

[0033] Preferably, the synchronous update of the triggering sequence based on the alignment in the phase change spectrum includes the following steps:

[0034] Based on the continuous recording of directional segments of current change sequences, temperature change sequences, and vibration change sequences using in-phase change sequences;

[0035] When the current change sequence exhibits continuous step changes under a unified time coordinate, the starting time segment of the load jumps again is determined. Within the starting time segment, the phase start points of the current change sequence, temperature change sequence, and vibration change sequence are re-extracted, and the overall time shift is performed according to the alignment in the phase change spectrum.

[0036] Based on the duration of the statistically consistent segments of the current change sequence, temperature change sequence, and vibration change sequence that have completed the overall time shift, an updated dynamic priority adjustment table is generated.

[0037] The trigger order is rearranged according to the updated dynamic priority adjustment table, and the alarm triggering conditions are executed according to the rearranged trigger order.

[0038] Preferably, the starting time segment of the load change is determined by the current change sequence maintaining a step change in the same direction for more than five consecutive time scales under a unified time coordinate; the directional segment statistics are based on the continuous unidirectional change segments of the current change sequence, temperature change sequence and vibration change sequence under a unified time coordinate.

[0039] The technical effects and advantages provided by this application in the above technical solution are as follows:

[0040] This application constructs a phase change map of power equipment during operation and compensates for phase shifts in current, temperature, and vibration change sequences. This allows change sequences with different response rhythms to form a co-phase change sequence on a unified time coordinate, fundamentally eliminating the trend misalignment problem caused by response lag during load surges. After regenerating the trigger sequence based on the co-phase change sequence, the alarm triggering logic of the power equipment can maintain consistency with the load change rhythm, thereby preventing abnormal features from being misinterpreted as a recovery state in the time dimension and improving the directional stability and continuity of power equipment anomaly identification.

[0041] This application uses a dynamic priority adjustment table to rearrange the decision path of power equipment in a temporal sequence, and writes the continuous interval of abnormal characteristics in the same phase change sequence into the trigger condition. Simultaneously, it excludes short-duration recovery segments from participating in the decision, ensuring that the abnormal trend of power equipment is continuously expressed on the time axis. When the load of the power equipment experiences another sudden change, the trigger sequence is updated synchronously using the alignment method in the phase change map. This ensures that the decision path of the power equipment remains stable and continuous under multiple load fluctuation scenarios, preventing dangerous states from being misidentified as normal fluctuations, thereby reducing the hidden dangers caused by the accumulation of risks during the operation of power equipment. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings.

[0043] Figure 1 This is a flowchart of the method in this application;

[0044] Figure 2 This is a flowchart of the method for constructing phase change maps based on each phase starting point in this application;

[0045] Figure 3 This application provides a flowchart of the method for generating a dynamic priority adjustment table based on phase consistency.

[0046] Figure 4 This is a flowchart of the method for synchronously updating the triggering order based on the alignment in the phase change spectrum. Detailed Implementation

[0047] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0048] This application provides, as follows: Figures 1 to 4 The method for analyzing all-element monitoring data of power equipment that integrates multiple rules, as shown, includes the following steps:

[0049] Step 1: Collect the current change sequence, temperature change sequence, and vibration change sequence of the power equipment during the load jump process. Unfold the current change sequence, temperature change sequence, and vibration change sequence under a unified time coordinate, extract the phase start point of each change sequence, and construct a phase change spectrum based on each phase start point.

[0050] The phase change spectrum is constructed based on each phase starting point. The specific steps are as follows:

[0051] When the power equipment is in continuous operation, current, temperature and vibration are continuously collected. The data collection process records the changes in values ​​according to a predetermined sampling period and assigns a corresponding collection time mark to each record. When a load jump occurs during operation, the starting time segment of the load jump is determined by continuously observing the time series of current records. When the current value shows a step change between adjacent time scales and the change remains in the same direction in several subsequent consecutive time scales.

[0052] After determining the starting time interval, complete data segments are extracted from the current records for the preset time before the load jump, the period during the load jump, and the preset time after the load jump. At the same time, data for the corresponding time periods are extracted from the temperature and vibration records according to the same time interval.

[0053] Subsequently, the original acquisition time stamps of the current change sequence, temperature change sequence, and vibration change sequence were uniformly mapped to the same continuous time axis, and the original scattered acquisition time scale was converted into a continuously increasing unified time coordinate scale, so that the three types of change sequences were arranged point by point on the same time axis, forming the current change sequence, temperature change sequence, and vibration change sequence under the unified time coordinate.

[0054] It should be noted that the criteria for determining a step change are defined as follows:

[0055] A step change is considered to have occurred when the difference between the current value at a certain time scale and the current value at the previous time scale exceeds a preset change threshold. This preset change threshold can be set based on the rated current value of the power equipment or the maximum fluctuation range within a historical stable operating range; for example, it can be a percentage of the rated current or a multiple of the maximum fluctuation range of the stable section. In addition to the value change between two adjacent time scales exceeding this threshold, the change direction must remain consistent for several consecutive time scales, i.e., continuously rising or continuously falling, and there should be no reverse change exceeding a preset small fluctuation range within this continuous period.

[0056] A step change in current value is determined only when both conditions are met simultaneously: "the magnitude of a single change exceeds the preset change threshold" and "the subsequent continuous time scale changes in the same direction". This serves as the basis for identifying the starting time segment of the load jump.

[0057] After completing the unified time coordinate expansion, the numerical change trajectory of the current change sequence under the unified time coordinate is continuously scanned to locate the first time scale when the current value enters a state of continuous rise or continuous fall from a stable fluctuation state, and this time scale is recorded as the phase start point of the current change sequence.

[0058] Next, under the same unified time coordinate, the temperature change sequence was observed point by point. When the temperature value changed from a small fluctuation to a continuous rising range and maintained the same direction of change in subsequent time scales, the corresponding time scale was recorded as the phase start point of the temperature change sequence. Subsequently, the vibration change sequence was analyzed under the same time coordinate. When the vibration amplitude changed from the normal fluctuation range to a range of continuous increase or decrease and maintained the same direction of change in subsequent time scales, the corresponding time scale was recorded as the phase start point of the vibration change sequence.

[0059] It should be noted that the judgment rules for the transitions between the three types of states—"from a stable fluctuation state to a sustained upward or downward state, from a small fluctuation state to a continuous upward phase, and from a normal fluctuation range to a continuous increase or decrease range"—can be uniformly defined according to the constraints of fluctuation amplitude, continuity, and direction consistency. Specifically:

[0060] Under a unified time coordinate, a fluctuation range is first set. When the numerical change of the current change sequence always falls within the fluctuation range and the direction of change is not fixed within a certain number of time scales, the segment is identified as a stable fluctuation state.

[0061] When a certain time scale begins, if the current value increases or decreases unidirectionally for multiple consecutive time scales, and the change amplitude between adjacent time scales continuously exceeds the fluctuation amplitude range, and no reverse change occurs within the continuous segment exceeding a preset small range, the segment is identified as a continuously rising or continuously falling state, and the starting time scale of the segment is taken as the state transition point.

[0062] The temperature change sequence and the vibration change sequence use the same judgment logic. The temperature change sequence is defined as a small fluctuation state by the range of smaller fluctuation amplitude, and the vibration change sequence is defined as a normal fluctuation range by the amplitude fluctuation range under normal operation. When the corresponding sequence maintains unidirectional change in multiple consecutive time scales and the change amplitude continuously exceeds the original fluctuation range, and the direction remains consistent, it is identified as entering a continuous rising segment, a continuous increase interval, or a continuous decrease interval, thereby determining the phase start point of each change sequence.

[0063] In the above process, the phase start points of the current change sequence, temperature change sequence, and vibration change sequence are all marked with specific time scales under a unified time coordinate, and maintain a one-to-one correspondence with the corresponding change sequence.

[0064] After obtaining the phase start points of the current change sequence, the temperature change sequence, and the vibration change sequence, the specific time scales of the three phase start points under a unified time coordinate are recorded side by side to form a phase record set containing the three time scale values.

[0065] Based on the order of the three time scales, the sequential relationship of the responses of the current change sequence, temperature change sequence, and vibration change sequence during the load jump is determined, and the time interval between two adjacent phase start points is recorded. On a unified time coordinate, with each phase start point as the center, the time scale is traced back to the starting time scale when the current change sequence was in the stable range before the load jump, and then extended backward to the time scale after the vibration change sequence completed the continuous change segment, thus forming a complete change range around the phase start point for each change sequence.

[0066] By labeling the complete variation intervals of the three variation sequences side by side on a unified time coordinate, a time correspondence between the three types of variation sequences is constructed, enabling the current variation sequence, temperature variation sequence, and vibration variation sequence to form a directly comparable response rhythm relationship under a unified time coordinate.

[0067] "On a unified time coordinate, taking the starting point of each phase as the center, we trace back to the starting time scale when the current change sequence before the load jump is in the stable range, and extend backward to the time scale after the vibration change sequence completes the continuous change segment, thus forming a complete change range around the starting point of each change sequence." The following is a detailed explanation of this part:

[0068] After determining the phase start points of the current change sequence, temperature change sequence, and vibration change sequence, we do not simply record this "start time point" alone, but rather extend forward and backward around this time point to determine a complete time range. The time range covers the stable phase before the load jump, the jump occurrence phase, and the entire process of the gradual completion of various responses after the jump. The complete time range is used to represent a complete dynamic response process.

[0069] In this way, each phase start point is no longer just an isolated point in time, but corresponds to a complete time period covering the steady state before the load jump, the jump response process, and the end of the response phase. This time period reflects the overall dynamic change process caused by a load jump, providing a unified time range basis for subsequent alignment and comparison.

[0070] Finally, under a unified time coordinate, the phase start points of the current change sequence, the temperature change sequence, and the vibration change sequence are marked in time scale order, and the position markers corresponding to the three phase start points are plotted on a unified time axis. At the same time, the numerical change trajectories of the current change sequence, temperature change sequence, and vibration change sequence before and after the phase start points are preserved. Using the unified time coordinate as the horizontal reference, the time interval relationship between the three phase start points is fixedly recorded and arranged on the time axis to form an arrangement structure reflecting the order of response, thus forming a complete phase change spectrum.

[0071] This phase change map not only includes the phase start points of the current change sequence, temperature change sequence, and vibration change sequence, but also the time correspondence and change range of the three change sequences throughout the entire load jump process, providing a clear time reference for subsequent phase offset compensation processing based on the phase start point.

[0072] Step 2: Based on the phase start point in the phase change spectrum, perform phase shift compensation processing on the current change sequence, temperature change sequence and vibration change sequence respectively, align the phase start points to obtain the same phase change sequence.

[0073] Phase shift compensation processing was performed on the current change sequence, temperature change sequence, and vibration change sequence respectively to obtain in-phase change sequences. The specific steps are as follows:

[0074] Based on the established phase change spectrum, the specific time scale value corresponding to the phase start point of the current change sequence in a unified time coordinate is read, the specific time scale value corresponding to the phase start point of the temperature change sequence in a unified time coordinate is read, and the specific time scale value corresponding to the phase start point of the vibration change sequence in a unified time coordinate is read. The three time scale values ​​are arranged in chronological order to form a time scale set.

[0075] The earliest time scale in the time scale set is selected as the unified alignment reference time scale, and this unified alignment reference time scale is determined as the unified reference point for subsequent phase offset compensation processing. Then, the time interval between the phase start point time scale of the current change sequence and the unified alignment reference time scale is calculated, the time interval between the phase start point time scale of the temperature change sequence and the unified alignment reference time scale is calculated, and the time interval between the phase start point time scale of the vibration change sequence and the unified alignment reference time scale is calculated. The three time interval values ​​are recorded as the time offset of the current change sequence, the time offset of the temperature change sequence, and the time offset of the vibration change sequence, respectively.

[0076] While recording the time offset, the complete arrangement of each change sequence in the phase change spectrum under a unified time coordinate is preserved, so that the original time order and time interval relationship of each change sequence remain unchanged.

[0077] After completing the time offset recording, the entire time scale of the current change sequence is translated under a unified time coordinate. The data points corresponding to each time scale in the current change sequence are moved synchronously according to the time offset of the current change sequence, so that the phase start point of the current change sequence coincides with the unified alignment reference time scale after the movement. During the movement, the time interval and numerical order between the data points in the current change sequence are kept unchanged, and only their overall position in the unified time coordinate is adjusted.

[0078] After completing the translation process of the current change sequence, the same overall translation process is performed on all time scales of the temperature change sequence under a unified time coordinate. The data points corresponding to each time scale in the temperature change sequence are moved synchronously according to the time offset of the temperature change sequence, so that the phase start point of the temperature change sequence coincides with the unified alignment reference time scale, while keeping the time interval relationship and numerical arrangement order within the temperature change sequence unchanged.

[0079] Subsequently, the vibration change sequence is subjected to the same overall translation process on all time scales under a unified time coordinate. The data points corresponding to each time scale in the vibration change sequence are moved synchronously according to the time offset of the vibration change sequence, so that the phase start point of the vibration change sequence coincides with the unified alignment reference time scale, and the time interval and numerical change order between data within the vibration change sequence remain unchanged.

[0080] After the above translation process, the current change sequence, temperature change sequence, and vibration change sequence are arranged in a unified time coordinate with a common starting point of a unified alignment reference time scale.

[0081] After the current change sequence, temperature change sequence, and vibration change sequence have been shifted over time, the three shifted change sequences are superimposed and arranged according to a unified time coordinate, so that each time scale of the current change sequence, temperature change sequence, and vibration change sequence can be compared accordingly.

[0082] In a unified time coordinate system, the phase start points of the current change sequence, the temperature change sequence, and the vibration change sequence are all located at a unified alignment reference time scale position, and the continuous transition relationship between the historical segments before the phase start point and the change segments after the phase start point of the three change sequences is maintained.

[0083] After unifying and aligning the reference time scale, the subsequent change trajectories of the current change sequence, temperature change sequence, and vibration change sequence are simultaneously unfolded according to the unified time coordinate sequence, so that the three change sequences extend backward from the same time starting point in the response stage after the load jump, thus forming a change sequence in phase.

[0084] Through the above continuous processing, the current change sequence, temperature change sequence, and vibration change sequence are unfolded under a unified time coordinate with the same phase starting point as a reference, eliminating the time misalignment caused by response time differences, and providing a consistent time basis for subsequent adjustment of the triggering order around the same phase change sequence.

[0085] Step 3: Calculate the phase consistency between different phase change sequences, generate a dynamic priority adjustment table based on the phase consistency, and reorder the original alarm triggering sequence according to the dynamic priority adjustment table to form a triggering sequence corresponding to the load change rhythm.

[0086] A dynamic priority adjustment table is generated based on the phase consistency level. The specific steps are as follows:

[0087] Based on the obtained in-phase change sequence, starting from the unified alignment reference time scale, a continuous time segment covering the entire process of load jump is selected. For example, starting from the unified alignment reference time scale, all time scales within the range of 30 to 60 seconds are continuously selected. The data corresponding to each time scale of the current change sequence, temperature change sequence, and vibration change sequence in this continuous time segment are expanded point by point and arranged in parallel according to the chronological order.

[0088] Based on the parallel recording, the direction of change of the current change sequence at each time scale is marked. When the current time scale value is greater than the previous time scale value, it is marked as an upward change. When the current time scale value is less than the previous time scale value, it is marked as a downward change. When the change amplitude of three consecutive time scale values ​​is within a preset small range, it is marked as a stable change. Subsequently, the direction of the temperature change sequence and the vibration change sequence is marked point by point in the same way, forming a combination of direction marks for the three types of change sequences at the same time scale.

[0089] It should be noted that the preset small range is set based on the natural fluctuation characteristics of the change sequence during the stable operation phase, specifically as follows:

[0090] Before a sudden load jump occurs, select a time interval in which the current change sequence, temperature change sequence, or vibration change sequence is in a stable operating state, and statistically analyze the numerical change amplitude between adjacent time scales within this time interval. Then, determine an amplitude range that represents the normal fluctuation level, for example, taking the maximum or average change amplitude within this interval as a reference value.

[0091] Based on this, the reference value is used as the upper limit of the "preset small range". When the value change between the next three consecutive time scales does not exceed this range, it is considered to be in a stable change state. Thus, the time segment is marked as a stable change segment to distinguish it from rising or falling change segments.

[0092] After marking the direction of all time scales, the time scales that maintain the same direction mark are merged into a continuous direction segment, and the start and end time scales and duration of each continuous direction segment are recorded, so as to obtain the distribution of direction segments of current change sequence, temperature change sequence and vibration change sequence under a unified time coordinate.

[0093] After completing the directional segment distribution record, using a unified time coordinate as a reference, each continuous directional segment of the current change sequence is compared with the directional segments of the temperature change sequence within the same time scale range. When the directional markings of the two change sequences are consistent within the same time segment and the duration exceeds the preset time length, the time segment is recorded as a phase-consistent segment, and the duration of the phase-consistent segment is recorded.

[0094] The vibration change sequence is then compared within the same time interval. When the direction markings of the current change sequence, temperature change sequence, and vibration change sequence are completely consistent within the same time interval and the duration exceeds the preset time length, the time interval is recorded as the three-sequence phase-consistent interval, and its duration and occurrence location are recorded.

[0095] It should be noted that the preset time length is set based on the typical duration characteristics of the change sequence during a sudden load jump, specifically:

[0096] Multiple load jump processes were selected from historical operating data. The time intervals in which the current change sequence, temperature change sequence, and vibration change sequence continuously changed in the same direction were statistically analyzed. The smallest continuous time interval that could stably reflect the rhythm of load change was extracted as a reference. Based on this, the number of time scales corresponding to the smallest continuous time interval was set as the "preset time length" to distinguish between short-term fluctuations and real synchronous change processes.

[0097] When three changing sequences have the same direction markings and the duration reaches or exceeds the preset time length within the same time interval, they are identified as having synchronous change characteristics and are thus recorded as phase-consistent segments.

[0098] After completing the segment-by-segment comparison of all time intervals, the number of phase-consistent segments between the current change sequence and the temperature change sequence, the number of phase-consistent segments between the current change sequence and the vibration change sequence, and the number of phase-consistent segments between the temperature change sequence and the vibration change sequence are counted. The three statistical results are then arranged according to the sum of the durations to form a record of the phase consistency between each change sequence.

[0099] After obtaining the phase consistency records, the current change sequence, temperature change sequence, and vibration change sequence are treated as independent objects and their participation is ranked. Specifically, the cumulative duration of the current change sequence participating in the phase consistency segment of the three sequences is combined with the cumulative duration of the current change sequence participating in the phase consistency segment of the two sequences and recorded together, and compared with the corresponding records of the temperature change sequence and vibration change sequence. The three types of change sequences are arranged in descending order of cumulative duration, with the change sequence with the longest cumulative duration placed first, the change sequence with the second longest cumulative duration placed second, and the change sequence with the shortest cumulative duration placed last.

[0100] After the arrangement is completed, the arrangement order is used as the sorting result of the phase consistency during the current load jump process, and a dynamic priority adjustment table is generated according to the sorting result. The priority order of the current change sequence, temperature change sequence and vibration change sequence and the corresponding cumulative duration data are clearly recorded in the dynamic priority adjustment table, and the time segment records under the unified time coordinate are retained as additional explanations.

[0101] After generating the dynamic priority adjustment table, each alarm condition in the original alarm triggering sequence is listed according to the current change sequence alarm condition, temperature change sequence alarm condition, and vibration change sequence alarm condition, and mapped to the priority order in the dynamic priority adjustment table.

[0102] It should be noted that the original alarm triggering order is usually derived from the rule configuration order pre-set in the power equipment monitoring or operation system, which can be obtained by reading the existing alarm rule library or judgment logic configuration. In practical applications, various alarm conditions are fixedly sorted according to empirical rules, equipment operation mechanisms or safety strategies when the system is deployed. For example, they are arranged in a way that alarms related to current change sequence take priority, followed by temperature change sequence, and then vibration change sequence. During operation, the alarms are triggered and judged in this predetermined order.

[0103] This order is usually stored in the form of a rule list, decision chain, or triggering process. Before dynamic priority adjustment, the original alarm triggering order can be obtained by reading this established arrangement structure, which can then serve as a reference for subsequent priority mapping and reordering.

[0104] If the current change sequence is listed first in the dynamic priority adjustment table, then the alarm condition corresponding to the current change sequence in the original alarm triggering sequence will be moved to the first position of the triggering sequence, the alarm condition corresponding to the change sequence listed second will be placed in the second position, and the alarm condition corresponding to the change sequence listed third will be placed in the third position, thus forming a new alarm triggering sequence.

[0105] During the reordering process, the judgment content and trigger threshold of each alarm condition remain unchanged. Only their order in the triggering sequence is adjusted so that the new triggering sequence is consistent with the sorting result of the phase consistency of each change sequence in the same phase change sequence.

[0106] Through the above reordering process, a triggering sequence corresponding to the load change rhythm is formed, so that the triggering sequence is triggered in descending order of phase consistency throughout the entire load jump process, thereby ensuring that the alarm triggering logic and the rhythm relationship of the same phase change sequence remain consistent.

[0107] Step 4: Based on the dynamic priority adjustment table, the current decision path is rearranged in time sequence, the continuous interval of abnormal features in the same phase change sequence is written into the trigger condition, and short instantaneous recovery segments are excluded from the decision path.

[0108] Based on the dynamic priority adjustment table, the current decision path is rearranged in time sequence. The specific steps are as follows:

[0109] After the dynamic priority adjustment table has been formed and the triggering order has been obtained according to the priority order, each alarm triggering condition in the original judgment path is listed one by one and reordered according to the order in the dynamic priority adjustment table. Specifically, if the first priority in the dynamic priority adjustment table corresponds to the current change sequence, the alarm triggering condition related to the current change sequence in the original judgment path is moved to the beginning position of the judgment path, the alarm triggering condition related to the change sequence corresponding to the second priority is moved to the second position, and the alarm triggering condition related to the change sequence corresponding to the third priority is moved to the third position, thus forming a new execution order.

[0110] After rearrangement, the new execution order is matched segment by segment with the unfolded state of the in-phase change sequence under a unified time coordinate, so that each trigger condition in the decision path corresponds to a specific time segment in the in-phase change sequence. Then, a time interval of several consecutive seconds from the unified alignment reference time scale is selected in the unified time coordinate, and the abnormal segments of the current change sequence, temperature change sequence, and vibration change sequence within the time interval are marked segment by segment. The start time scale, end time scale, and duration of each abnormal segment are recorded, and the abnormal segment is matched one by one with the reordered trigger conditions, thereby completing the decision path timing rearrangement based on the dynamic priority adjustment table.

[0111] It should be noted that the abnormal segment refers to the time interval in the current change sequence, temperature change sequence, or vibration change sequence that deviates from the normal operating fluctuation range and continuously exhibits unidirectional abnormal changes under a unified time coordinate. Specifically, it is manifested in that within multiple consecutive time scales, the direction of numerical change remains consistent and the magnitude of change exceeds the fluctuation range of the corresponding sequence during the stable operating phase.

[0112] Abnormal sections typically correspond to abnormal response phases that occur in power equipment during load surges, such as continuous current increases or decreases, continuous temperature increases, and continuous expansion of vibration amplitude. By recording the start time, end time, and duration of these abnormal sections, this continuous change process is considered a complete abnormal section and used to correspond with the alarm triggering conditions in the judgment path.

[0113] After the timing reordering of the judgment path is completed, the duration interval of the abnormal feature in the same phase change sequence is written into the trigger condition. Specifically, the execution rule of each trigger condition is to add a limitation on the duration length. For example, the following constraint is added to the trigger condition corresponding to the current change sequence: when the current change sequence maintains a unidirectional upward state for ten consecutive time scales from a certain time scale under a unified time coordinate, and the direction of numerical change in this segment is consistent, the corresponding alarm condition is triggered.

[0114] Add the following constraints to the triggering conditions corresponding to the temperature change sequence:

[0115] When the temperature change sequence forms a continuous rising segment under a unified time coordinate and covers at least eight consecutive time scales, the corresponding alarm condition is triggered; the following constraint is added to the trigger condition corresponding to the vibration change sequence: when the vibration change sequence shows an amplitude expansion state on multiple consecutive time scales under a unified time coordinate and the duration exceeds the preset time length, the corresponding alarm condition is triggered.

[0116] By directly writing the duration of the abnormal feature in the same phase change sequence into the trigger condition, the establishment of the trigger condition depends not only on the instantaneous numerical state of a certain time scale, but also on the direction and duration of change within a continuous time segment, thereby ensuring the continuous expression of the abnormal feature in the time dimension.

[0117] After writing the duration of the abnormal feature into the trigger condition, the short instantaneous recovery segments appearing in the same phase change sequence are specifically identified and excluded from the judgment path. The specific process is to scan the direction marks of the current change sequence, temperature change sequence and vibration change sequence point by point under a unified time coordinate. When a reverse change segment with a duration of less than three time scales appears within the already determined abnormal duration interval, the reverse change segment is recorded as a short instantaneous recovery segment, and its original position in the unified time coordinate is kept unchanged.

[0118] During the execution of the decision path, when the scan reaches the time scale corresponding to the short-duration recovery segment, the recovery decision condition is not triggered, nor is the timing record of the current anomaly duration interval terminated. Instead, the time scale is continued to be counted within the original anomaly duration interval, keeping the anomaly segment continuous on the time axis. After completing the above processing, the decision path executes the trigger conditions item by item in the order determined by the dynamic priority adjustment table. During the execution process, it always tracks the continuous unfolding state of the in-phase change sequence under a unified time coordinate, so that the anomaly features form a complete continuous interval in the time dimension. At the same time, the short-duration recovery segment is excluded from the decision, thereby maintaining the temporal continuity of the decision path and the consistency of the execution order in the case of sudden load jumps.

[0119] Step 5: After the decision path is completed and the timing is rearranged, its evolution rhythm is continuously tracked based on the same phase change sequence. When the load changes abruptly again, the triggering order is updated synchronously according to the alignment method in the phase change spectrum to maintain the stability of the decision direction and prevent dangerous states from being identified as recovery states.

[0120] When the load changes abruptly again, the triggering sequence is updated synchronously according to the alignment in the phase change spectrum. The specific steps are as follows:

[0121] After the path is determined to be rearranged and a new triggering sequence is formed, the current change sequence, temperature change sequence and vibration change sequence are continuously extended and recorded in the same phase change sequence framework, using a unified time coordinate as a continuous time axis. Specifically, after aligning the reference time scale, the current value, temperature value and vibration value corresponding to each newly added time scale are registered point by point and arranged in chronological order in the unified time coordinate.

[0122] During continuous recording, the directional state of the current change sequence is marked on a scale. When the current time scale value is higher than the previous time scale value, it is marked as rising; when the current time scale value is lower than the previous time scale value, it is marked as falling; when the change amplitude of three consecutive time scale values ​​is within a preset small range, it is marked as stable. The temperature change sequence and the vibration change sequence are marked on a scale in the same way, and the directional marking results of the three change sequences are arranged in parallel under a unified time coordinate.

[0123] It should be noted that the preset micro-range is set based on the natural fluctuation range of power equipment during stable operation, specifically as follows:

[0124] Before a sudden load jump occurs, select a time interval in which the current change sequence, temperature change sequence, or vibration change sequence is in a stable operating state, and statistically analyze the numerical change amplitude between adjacent time scales within this interval, using the maximum or average value of this change amplitude as the reference limit.

[0125] Based on this, the reference limit is set as the range of the "preset small interval". When the numerical changes between three consecutive time scales fall within this range, it is determined to be in a stable change state, which is used to distinguish between continuously rising or continuously falling change segments.

[0126] During continuous recording, the newly formed continuous directional segments are connected end-to-end with the directional segments formed during the previous load jump, so that the in-phase change sequence forms an uninterrupted continuous rhythm chain in a unified time coordinate, thereby completing the continuous tracking of the evolution rhythm of the in-phase change sequence.

[0127] During continuous tracking, when the current change sequence shows a step increase or step decrease for more than five consecutive time scales under a unified time coordinate, and the direction of the change is different from the direction of the previous stable segment, it is determined as the starting signal of another sudden change in load.

[0128] After confirming a new sudden load change, the starting time scale is traced back to the starting time scale of the most recent stable segment, and then extended to the ending time scale of the continuously changing segment. The current change sequence, temperature change sequence, and vibration change sequence within this segment are re-analyzed. Subsequently, the phase start points of the current change sequence, temperature change sequence, and vibration change sequence are re-extracted within this segment, and the specific time scales of the three phase start points on a unified time coordinate are recorded.

[0129] The time scales of the three new phase start points are compared with the unified alignment reference time scales recorded in the existing phase change spectrum. The time intervals between them are calculated. The current change sequence, temperature change sequence and vibration change sequence are shifted in time according to the alignment method in the existing phase change spectrum. This ensures that the new phase start points fall in the same alignment position as the existing phase change spectrum in the unified time coordinate, so that multiple load jump changes maintain the same alignment structure on the time axis.

[0130] After completing the new phase start point alignment process, the updated current change sequence, temperature change sequence and vibration change sequence are re-incorporated into the same phase change sequence framework for directional segment statistics. The start time scale, end time scale and duration of the newly formed continuous directional segments in the unified time coordinate are recorded segment by segment. The number of directional consistent segments and the total duration between the three types of change sequences are recalculated according to the previously established statistical method.

[0131] Based on the new statistical results, the priority order of the current change sequence, temperature change sequence, and vibration change sequence is rearranged, a new dynamic priority adjustment table is generated, and this dynamic priority adjustment table is mapped to the current triggering order. The alarm conditions in the triggering order are rearranged according to the new priority order. After the triggering order is updated, the expression content of the triggering condition is kept unchanged in the continuous interval of the abnormal feature in the same phase change sequence, so that the new triggering order is executed sequentially according to the latest rhythm relationship under a unified time coordinate.

[0132] Through the above synchronous update process, the judgment path is rhythmically calibrated using the alignment method in the phase change spectrum every time the load changes abruptly, thereby maintaining the stability of the judgment direction and avoiding the misinterpretation of an abnormal state that has not yet ended as a recovery state in time during continuous load fluctuations.

[0133] This application constructs a phase change map of power equipment during operation and compensates for phase shifts in current, temperature, and vibration change sequences. This allows change sequences with different response rhythms to form a co-phase change sequence on a unified time coordinate, fundamentally eliminating the trend misalignment problem caused by response lag during load surges. After regenerating the trigger sequence based on the co-phase change sequence, the alarm triggering logic of the power equipment can maintain consistency with the load change rhythm, thus preventing abnormal features from being misinterpreted as recovery states in the time dimension and improving the directional stability and decision continuity of power equipment anomaly identification. This application uses a dynamic priority adjustment table to rearrange the decision path of the power equipment in time sequence and writes the duration interval of abnormal features in the co-phase change sequence into the trigger conditions, while excluding short-duration recovery segments from participating in the decision, ensuring that the abnormal trend of the power equipment is continuously expressed on the time axis. When the load of the power equipment experiences another sudden change, the trigger sequence is updated synchronously using the alignment method in the phase change map, ensuring that the decision path of the power equipment remains stable and continuous under multiple load fluctuation scenarios, preventing dangerous states from being misidentified as normal fluctuations, thereby reducing the hidden dangers caused by the accumulation of risks during power equipment operation.

[0134] The foregoing has only described certain exemplary embodiments of this application by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of this application. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of this application.

Claims

1. A method for analyzing all elements of power equipment monitoring data that integrates multiple rules, characterized in that: Includes the following steps: The current change sequence, temperature change sequence, and vibration change sequence of power equipment during load jump are collected, unfolded under a unified time coordinate, and the phase start point of each change sequence is extracted. A phase change spectrum is constructed based on each phase start point. Based on the phase start point in the phase change spectrum, phase shift compensation processing is performed on the current change sequence, temperature change sequence and vibration change sequence respectively, and the phase start points are aligned to obtain the same phase change sequence. The phase consistency between each change sequence is calculated based on the phase consistency, a dynamic priority adjustment table is generated based on the phase consistency, and the original alarm triggering order is reordered according to the dynamic priority adjustment table to form a triggering order corresponding to the load change rhythm. The current decision path is time-series rearranged based on a dynamic priority adjustment table. The duration of abnormal features within the same-phase change sequence is written into the trigger condition, and short-instantaneous recovery segments are excluded from the decision path. Specifically, this includes: Based on the dynamic priority adjustment table, the alarm conditions for current change sequence, temperature change sequence, and vibration change sequence in the original judgment path are reordered, and the reordered trigger conditions are matched with the abnormal sections of the same phase change sequence under the same time coordinate. Mark the abnormal duration intervals of current change sequences, temperature change sequences, and vibration change sequences, and write the start time scale, end time scale, and duration of the abnormal duration interval into the corresponding trigger conditions. Identify and record reverse change segments within the abnormal duration interval that have a duration less than a preset time scale as short instantaneous recovery segments, and exclude these short instantaneous recovery segments from the judgment process during the judgment path execution. After the path is determined to be rearranged in time, its evolution rhythm is continuously tracked based on the same phase change sequence. When the load changes abruptly again, the triggering order is updated synchronously according to the alignment in the phase change map. Generating a dynamic priority adjustment table based on phase consistency includes the following steps: In the same-phase change sequence, a continuous time segment is selected starting from the unified aligned reference time scale. The change direction of the current change sequence, temperature change sequence and vibration change sequence is marked point by point and merged to form a continuous direction segment. Based on the continuous directional segments, segment by segment comparison is performed under a unified time coordinate to record the phase-consistent segments and durations of current change sequences, temperature change sequences and vibration change sequences within the same time segment, forming a phase consistency record result. Based on the results of the phase consistency record, the cumulative duration of each change sequence participating in the phase consistency segment is calculated and arranged in order of cumulative duration to generate a dynamic priority adjustment table.

2. The method for analyzing all elements of power equipment monitoring data by integrating multiple rules as described in claim 1, characterized in that, Constructing a phase change map involves the following steps: When the current value undergoes a step change between adjacent time scales and maintains the same direction of change within a continuous time scale, the load jump start time segment is determined, and the current change sequence, temperature change sequence, and vibration change sequence within the corresponding load jump start time segment are extracted. Arrange the current change sequence, temperature change sequence, and vibration change sequence point by point on a unified time coordinate. The numerical change trajectories of the current change sequence, temperature change sequence, and vibration change sequence are scanned respectively, the phase start point is recorded, and a phase record set containing three time scale values ​​is formed. Based on the phase record set, the phase start points and corresponding change intervals of the current change sequence, temperature change sequence, and vibration change sequence are marked to form a phase change spectrum.

3. The method for analyzing all elements of power equipment monitoring data by integrating multiple rules as described in claim 1, characterized in that, Obtaining the in-phase transition sequence includes the following steps: Read the time scale values ​​of the phase start points of the current change sequence, temperature change sequence and vibration change sequence in the phase change spectrum on the unified time coordinate, calculate the time interval between the time scale of each phase start point and the unified alignment reference time scale and record it as the time offset. Based on the time offset, the current change sequence, temperature change sequence and vibration change sequence are all translated on the time scale under the unified time coordinate, and the phase starting point of each change sequence is moved to the unified alignment reference time scale position. The current change sequence, temperature change sequence, and vibration change sequence that have completed the overall translation process are superimposed and arranged on a unified time coordinate. The change trajectories of each change sequence are synchronously unfolded with a unified alignment reference time scale as a common starting point, forming a change sequence in phase.

4. The method for analyzing all elements of power equipment monitoring data by integrating multiple rules according to claim 3, characterized in that, The unified alignment reference time scale is selected as the earliest time scale among the phase start point time scales of the current change sequence, temperature change sequence, and vibration change sequence; the time offset is determined based on the time interval between each phase start point time scale and the unified alignment reference time scale; the same phase change sequence is judged by the coincidence of the phase start point under the unified time coordinate.

5. The method for analyzing all elements of power equipment monitoring data by integrating multiple rules according to claim 1, characterized in that, The phase consistency record results are sorted based on the cumulative duration of the directional consistent segments of the current change sequence, temperature change sequence, and vibration change sequence under a unified time coordinate; the dynamic priority adjustment table records the priority order of each change sequence and the corresponding time segment position; the triggering order is arranged according to the priority order in the dynamic priority adjustment table.

6. The method for analyzing all elements of power equipment monitoring data by integrating multiple rules according to claim 1, characterized in that, The abnormal duration interval is determined based on the number of time scales that change continuously in one direction under a unified time coordinate. The determination path is executed according to the triggering order in the dynamic priority adjustment table and in combination with the abnormal duration interval.

7. The method for analyzing all elements of power equipment monitoring data by integrating multiple rules according to claim 1, characterized in that, The steps involved in synchronously updating the triggering sequence based on the alignment method in the phase change map are as follows: Based on the continuous recording of directional segments of current change sequences, temperature change sequences, and vibration change sequences using in-phase change sequences; When the current change sequence exhibits continuous step changes under a unified time coordinate, the starting time segment of the load jumps again is determined. Within the starting time segment, the phase start points of the current change sequence, temperature change sequence, and vibration change sequence are re-extracted, and the overall time shift is performed according to the alignment in the phase change spectrum. Based on the duration of the statistically consistent segments of the current change sequence, temperature change sequence, and vibration change sequence that have completed the overall time shift, an updated dynamic priority adjustment table is generated. The trigger order is rearranged according to the updated dynamic priority adjustment table, and the alarm triggering conditions are executed according to the rearranged trigger order.

8. The method for analyzing all elements of power equipment monitoring data by integrating multiple rules according to claim 7, characterized in that, The starting time segment of the load change is determined by the current change sequence maintaining a step change in the same direction for more than five consecutive time scales under a unified time coordinate; the directional segment statistics are based on the continuous unidirectional change segments of the current change sequence, temperature change sequence and vibration change sequence under a unified time coordinate.

Images