A sensor-based motor energy efficiency real-time monitoring system

Abstract

The application discloses a kind of motor energy efficiency real-time monitoring systems based on sensor, it is related to motor monitoring technical field, in quasi-steady interval is analyzed, calculates comprehensive efficiency output energy efficiency evaluation dataset, combines temperature dataset, vibration characteristic dataset and harmonic current spectrum dataset analysis energy efficiency deviation source, output loss decomposition result, compare current operating parameter with the best efficiency point in baseline model dataset, output efficiency deviation degree dataset, the present application realizes the real-time perception of energy efficiency state by collecting multi-source data, establishes baseline model dataset and self-adapting update, can be automatically corrected with the aging of equipment and working condition change, carries out energy efficiency decomposition under quasi-steady condition, outputs copper loss, iron loss and mechanical loss result, realizes energy efficiency deviation explainability, completes abnormal attribution and credibility score in combination with multi-channel consistency analysis, reduces false alarm and improves diagnostic reliability, to realize the dynamic monitoring and intelligent optimization of motor energy efficiency.

Description

Technical Field

[0001] This invention relates to the field of motor monitoring technology, and in particular to a sensor-based real-time motor energy efficiency monitoring system. Background Technology

[0002] With the widespread use of motors in industrial production, water supply systems, HVAC, energy transmission and other fields, motor energy efficiency has become a core indicator affecting system energy consumption and operating costs. In existing technologies, energy efficiency monitoring mostly relies on single electrical measurement methods, which can only provide indicators such as voltage, current or power factor, and cannot reflect the true energy efficiency level of motors under different loads, hydraulic coupling and mechanical conditions.

[0003] Currently, Chinese invention patent application number CN202211381584.2 discloses a method, system, electronic device, and storage medium for real-time energy efficiency monitoring of reciprocating compressor units. The method includes the following steps: A. Constructing a calculation model for the output power of the drive motor, calculating the motor output power using the motor's rated voltage, actual operating current, motor power factor, and efficiency under actual load conditions; B. Constructing a calculation model for the total compression power of the reciprocating compressor, calculating the total compression power using the first-stage compression indication power and the second-stage compression indication power; during the calculation of the two-stage compression indication power, correcting the adiabatic index of the compressed gas in the two stages using the actual temperatures of the compressor's inlet and outlet; C. Calculating the total efficiency of the reciprocating compressor using the ratio of the total compression power of the reciprocating compressor to the output power of the drive motor, and completing real-time energy efficiency monitoring of the reciprocating compressor unit based on the total efficiency index of the reciprocating compressor.

[0004] The aforementioned technologies are insufficient for real-time fusion analysis of multi-source data on motors under complex operating conditions and for interpretable identification of energy efficiency deviations, resulting in low accuracy of energy efficiency monitoring results and a lack of dynamic self-correction capabilities. Summary of the Invention

[0005] The technical problem solved by this invention is that the existing technology is difficult to achieve real-time fusion analysis of multi-source data of motors under complex operating conditions and interpretable identification of energy efficiency deviations, resulting in low accuracy of energy efficiency monitoring results and a lack of dynamic self-correction capability.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] A sensor-based real-time monitoring system for motor energy efficiency includes a data acquisition and alignment module, a baseline energy efficiency modeling module, an event identification and labeling module, an energy efficiency estimation and decomposition module, a source attribution assessment module, and an online calibration and feedback module.

[0008] The data acquisition and alignment module acquires electrical parameter datasets, temperature datasets, mechanical operation datasets, hydraulic parameter datasets, heat exchange datasets, and control and environmental datasets, and synchronizes them with the system control clock as a unified time reference. It then performs windowing and feature alignment on the sampling sequence and outputs an aligned dataset.

[0009] The baseline energy efficiency modeling module detects changes in flow rate, pressure difference, and rotational speed based on the aligned dataset. When the change is within a stable threshold, it is determined to be in a steady-state operating range. The corresponding data is extracted to establish an energy efficiency curve and generate a baseline model dataset.

[0010] The event labeling module compares the rate of change of flow rate, valve opening and rotation speed. When it is in a stable state, it outputs a quasi-steady-state dataset. When it detects start-up, shutdown, sudden pressure change or rapid valve adjustment, it outputs an event labeling dataset.

[0011] The energy efficiency estimation decomposition module calculates the comprehensive efficiency and outputs the energy efficiency assessment dataset in the quasi-steady-state range. It analyzes the sources of energy efficiency deviation by combining temperature, vibration and harmonic characteristics, and outputs the loss decomposition results and efficiency deviation dataset.

[0012] The source attribution assessment module identifies energy efficiency deviation sources based on loss decomposition results and generates an anomaly source determination and credibility score dataset.

[0013] The online calibration and recharge module corrects the baseline model dataset based on the quasi-steady-state operating performance.

[0014] The energy efficiency estimation decomposition module performs a consistency analysis on the correspondence between electrical energy input and hydraulic output within the quasi-steady-state interval, and extracts the active power and flow data at the corresponding time from the aligned dataset.

[0015] By statistically analyzing the changing trend of the ratio of active power to flow rate within a time window, a dataset of unit transmission energy consumption is output.

[0016] The overall efficiency is calculated based on the matching relationship between flow rate and pressure difference, and the energy efficiency assessment dataset is output.

[0017] Based on the multi-source operation characteristics, the physical sources of energy efficiency deviation are distinguished, and the harmonic current spectrum is combined with the changes in motor housing temperature and stator winding temperature to identify the changing trends of copper loss and iron loss caused by electromagnetic load.

[0018] The radial vibration acceleration and envelope spectrum characteristics are monitored. These characteristics are used to identify the trend of mechanical loss caused by mechanical friction, bearing wear, or hydraulic pulsation. By comparing the change amplitude of different loss sources within the same analysis window, copper loss distribution dataset, iron loss distribution dataset, and mechanical loss distribution dataset are output.

[0019] The copper loss distribution dataset, iron loss distribution dataset, and mechanical loss distribution dataset are merged and output as the loss decomposition result.

[0020] The current flow rate, differential pressure, and motor speed parameters are compared with the optimal efficiency point operating characteristics recorded in the baseline energy efficiency model. The deviation between the operating point and the optimal efficiency point is calculated, and the efficiency deviation dataset is output.

[0021] The source attribution assessment module identifies the specific sources of energy efficiency deviation in the loss decomposition results and calculates a credibility score based on multi-source evidence information.

[0022] Compare the copper loss distribution dataset, iron loss distribution dataset, and mechanical loss distribution dataset within the same analysis window, and calculate the degree of correspondence between each loss channel and the change in overall efficiency.

[0023] When the copper loss ratio increases significantly and is accompanied by an increase in the motor housing temperature or stator winding temperature, the output deviation is attributed to an increase in electromagnetic load or winding aging.

[0024] When the iron loss ratio increases and the low-frequency distortion in the harmonic current spectrum is enhanced, the output deviation is attributed to changes in magnetic flux distribution or unstable drive control.

[0025] When the mechanical loss ratio increases and characteristic peaks appear in the radial vibration acceleration or envelope spectrum characteristics, the output deviation is attributed to bearing wear, rotor imbalance or pump body hydraulic fluctuations.

[0026] The deviation attribution is correlated with the corresponding copper loss distribution dataset, iron loss distribution dataset, and mechanical loss distribution dataset to output an anomaly source determination dataset;

[0027] The loss decomposition results, temperature dataset, vibration feature dataset, and hydraulic parameter dataset are compared to analyze the consistency between information from different channels and quantify the reliability of the deviation analysis results.

[0028] When multiple data channels reflect the same energy efficiency deviation trend or point to the same source of deviation within the same time period, a high reliability score is output.

[0029] When there are differences between data from different channels or when the data is affected by noise, a low confidence score is output.

[0030] The scoring results are corrected by combining the running background in the event labeling dataset, distinguishing between transient disturbances and real anomalies, and outputting a credibility score dataset.

[0031] Preferably, the data acquisition and alignment module acquires three-phase voltage phasors, three-phase current phasors, power factor, active power, reactive power, and harmonic current spectrum, and outputs an electrical parameter dataset.

[0032] Collect the motor housing temperature and stator winding temperature, and output a temperature dataset.

[0033] Collect motor speed data and output a mechanical operation dataset.

[0034] Collect flow rate, differential pressure, and valve opening, and output a hydraulic parameter dataset.

[0035] Collect the supply water temperature and return water temperature, and output a heat exchange dataset.

[0036] Collect driver internal torque estimation, driver switching frequency, driver alarm status and ambient temperature, and output as a control and environment dataset;

[0037] The electrical parameter dataset, temperature dataset, mechanical operation dataset, hydraulic parameter dataset, heat exchange dataset, and control and environment dataset are merged and output as the original measurement dataset;

[0038] Axial vibration acceleration and radial vibration acceleration are obtained, and the characteristic energy patterns of axial vibration acceleration and radial vibration acceleration are identified by envelope spectrum calculation method to form a vibration feature dataset. The characteristic energy patterns include axial vibration acceleration and radial vibration acceleration corresponding to mechanical wear, bearing damage or assembly eccentricity.

[0039] Using the system control clock as a unified time reference, the sampling sequences of the original measurement dataset and vibration feature dataset are divided into windows to establish the first correspondence between the electrical parameter dataset, vibration feature dataset and hydraulic parameter dataset, and the output is an aligned dataset.

[0040] Preferably, the baseline energy efficiency modeling module detects the changes in flow rate, differential pressure and motor speed in the aligned data within a preset time window. When the changes in flow rate, differential pressure and motor speed are all within a set stable range, the corresponding time period is determined to be a steady-state operating range.

[0041] Within the steady-state operating range, active power, reactive power, and hydraulic parameter datasets are extracted from the aligned dataset. The correspondence between the active power, reactive power, and hydraulic parameter datasets is recorded, and the steady-state operating dataset is output.

[0042] The steady-state operation dataset is correlated and calculated to output an energy efficiency curve dataset, which reflects the relationship between water delivery volume and electricity consumption.

[0043] The energy efficiency curve dataset and the steady-state operation dataset are used together to construct the baseline model dataset.

[0044] Preferably, the event identification and labeling module analyzes the changes in flow rate, valve opening, and motor speed over a continuous time period to determine whether the current operating state is in a quasi-steady state.

[0045] Within each monitoring window, the rate of change of flow rate, the rate of change of valve opening, and the rate of change of motor speed are calculated, and the magnitude of the changes in the rate of change of flow rate, the rate of change of valve opening, and the rate of change of motor speed are compared with the preset stable threshold.

[0046] When the rate of change of flow rate, the rate of change of valve opening, and the rate of change of motor speed are all within the allowable range and no rapid disturbance occurs, the current operating state is determined to be stable, and a quasi-steady-state dataset is output.

[0047] When the rate of change of flow rate, the rate of change of valve opening and the rate of change of motor speed fluctuate briefly but quickly stabilize in the subsequent period, the corresponding period is identified as a quasi-steady state with acceptable disturbance.

[0048] When a sudden change in motor speed, a rapid adjustment in valve opening within a short period of time, a sudden fluctuation in differential pressure, or the triggering of a drive alarm is detected, the corresponding time segment is identified as an event segment and an event-labeled dataset is output. The event-labeled dataset contains information on event type, duration, and direction of change of associated parameters.

[0049] Preferably, the online calibration and recharge module dynamically corrects the baseline model dataset based on the energy efficiency performance under quasi-steady-state operating conditions during continuous equipment operation;

[0050] Active power, flow rate, pressure difference and motor speed data are extracted from the aligned dataset and compared with the reference operating point in the baseline model dataset.

[0051] When the difference is within the normal fluctuation range, record the steady-state point distribution and output the steady-state verification dataset;

[0052] When the difference continues to shift but does not reach the abnormal limit, the energy efficiency reference value of the corresponding running point is adjusted in small steps according to the direction and magnitude of the shift, and the updated baseline model dataset is output.

[0053] After equipment overhaul, maintenance, or component replacement, representative operating points are extracted from the stable interval as strongly labeled data. The strongly labeled data is used as a high-confidence reference to correct the feature values ​​of the corresponding segment in the baseline model, and the corrected model dataset is output.

[0054] For each model revision, a version identifier is generated and the revision time, basis, and applicable scope are recorded, and the model version record dataset is output.

[0055] Preferably, the energy efficiency estimation decomposition module dynamically adjusts the weights based on the changing trends of the temperature dataset, vibration feature dataset, and harmonic current spectrum dataset, and allocates the weight ratios of copper loss, iron loss, and mechanical loss in real time according to the fluctuation amplitude and stability of various features within the same time window.

[0056] Preferably, the event identification and labeling module has an adaptive window update mechanism to monitor the change error of the flow dataset and the motor speed dataset in real time. When the change error of the two is detected to be continuously deviating from the threshold, the window is shortened; when the running state is stable, the window is extended, and the output window is adjusted to record the dataset.

[0057] Preferably, the source attribution assessment module performs a lag compensation process when generating the credibility score, and corrects the change step size of the credibility score by analyzing the difference between the event occurrence time and the energy efficiency recovery time.

[0058] The beneficial effects of this invention are as follows: By collecting multi-source data, this invention achieves real-time perception of energy efficiency status, establishes a baseline model dataset and adaptively updates it, which can be automatically corrected as equipment ages and operating conditions change. Under quasi-steady-state conditions, it performs energy efficiency decomposition and outputs copper loss, iron loss and mechanical loss results, making energy efficiency deviations interpretable. Combined with multi-channel consistency analysis, it completes anomaly attribution and credibility scoring, reduces false alarms and improves diagnostic reliability, thereby realizing dynamic monitoring and intelligent optimization of motor energy efficiency. Attached Figure Description

[0059] Figure 1 This is a basic flowchart of a sensor-based real-time monitoring system for motor energy efficiency, provided as an embodiment of the present invention. Detailed Implementation

[0060] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0061] Example, refer to Figure 1 This paper presents a sensor-based real-time monitoring system for motor energy efficiency, including a data acquisition and alignment module, a baseline energy efficiency modeling module, an event identification and labeling module, an energy efficiency estimation and decomposition module, a source attribution assessment module, and an online calibration and feedback module.

[0062] The data acquisition and alignment module acquires electrical parameter datasets, temperature datasets, mechanical operation datasets, hydraulic parameter datasets, heat exchange datasets, and control and environmental datasets, and synchronizes them with the system control clock as a unified time reference. It then performs windowing and feature alignment on the sampled sequences and outputs the aligned dataset.

[0063] The baseline energy efficiency modeling module detects changes in flow rate, pressure difference, and rotational speed based on the aligned dataset. When the change is within a stable threshold, it is determined to be in a steady-state operating range. The corresponding data is extracted to establish an energy efficiency curve and generate a baseline model dataset.

[0064] The event labeling module compares the rate of change of flow rate, valve opening, and rotation speed. When the flow rate is stable, it outputs a quasi-steady-state dataset. When a start-up, shutdown, differential pressure change, or rapid valve adjustment is detected, it outputs an event labeling dataset.

[0065] The energy efficiency estimation decomposition module calculates the overall efficiency within the quasi-steady-state range and outputs an energy efficiency assessment dataset. It analyzes the sources of energy efficiency deviation by combining temperature, vibration, and harmonic characteristics, and outputs loss decomposition results and efficiency deviation datasets.

[0066] The source attribution assessment module identifies sources of energy efficiency deviation based on loss decomposition results and generates anomaly source determination and credibility score datasets.

[0067] The online calibration and recharge module corrects the baseline model dataset based on the quasi-steady-state operating performance.

[0068] This invention achieves real-time perception of energy efficiency status by collecting multi-source data, establishes a baseline model dataset and adaptively updates it, which can automatically correct itself as equipment ages and operating conditions change. Under quasi-steady-state conditions, it performs energy efficiency decomposition and outputs copper loss, iron loss and mechanical loss results, making energy efficiency deviations interpretable. Combined with multi-channel consistency analysis, it completes anomaly attribution and credibility scoring, reduces false alarms and improves diagnostic reliability, thereby realizing dynamic monitoring and intelligent optimization of motor energy efficiency.

[0069] The data acquisition and alignment module collects three-phase voltage phasors, three-phase current phasors, power factor, active power, reactive power, and harmonic current spectrum, and outputs an electrical parameter dataset.

[0070] Collect the temperature of the motor housing and the temperature of the stator winding, and output a temperature dataset.

[0071] Collect motor speed data and output a mechanical operation dataset.

[0072] Collect flow rate, differential pressure, and valve opening, and output a hydraulic parameter dataset.

[0073] Collect the supply water temperature and return water temperature, and output a heat exchange dataset.

[0074] The system collects data on the driver's internal torque estimation, driver switching frequency, driver alarm status, and ambient temperature, and outputs a control and environmental dataset.

[0075] The electrical parameter dataset, temperature dataset, mechanical operation dataset, hydraulic parameter dataset, heat exchange dataset, and control and environment dataset are merged and output as the original measurement dataset.

[0076] Axial and radial vibration accelerations are acquired, and characteristic energy patterns of axial and radial vibration accelerations are identified by envelope spectrum calculation methods to form a vibration characteristic dataset. The characteristic energy patterns include axial and radial vibration accelerations corresponding to mechanical wear, bearing damage, or assembly eccentricity.

[0077] Using the system control clock as a unified time reference, the sampling sequences of the original measurement dataset and vibration feature dataset are divided into windows to establish the first correspondence between the electrical parameter dataset, vibration feature dataset and hydraulic parameter dataset, and the output is an aligned dataset.

[0078] The data acquisition and alignment module enables unified acquisition and time synchronization of multi-source signals during motor operation. By acquiring data on electrical parameters, temperature, mechanical properties, hydraulic parameters, heat exchange, and control environment, a complete raw measurement dataset is formed. This dataset is then combined with vibration characteristic data to monitor multiple physical quantities. The module performs time alignment based on the system control clock, establishing the correspondence between electrical, vibration, and hydraulic characteristics, and outputs an aligned dataset. This ensures the consistency of different types of sensor signals in terms of time and operating conditions, providing a precise data foundation for subsequent energy efficiency calculations, loss decomposition, and anomaly analysis.

[0079] The baseline energy efficiency modeling module detects the changes in flow rate, differential pressure, and motor speed in the aligned data within a preset time window. When the changes in flow rate, differential pressure, and motor speed are all within the set stable range, the corresponding time period is determined to be the steady-state operating range.

[0080] Within the steady-state operating range, active power, reactive power, and hydraulic parameter datasets are extracted from the aligned dataset. The correspondence between active power, reactive power, and hydraulic parameter datasets is recorded, and the steady-state operating dataset is output.

[0081] The steady-state operation dataset is correlated and calculated to output an energy efficiency curve dataset, which reflects the relationship between water delivery volume and electricity consumption.

[0082] The energy efficiency curve dataset and the steady-state operation dataset are used together to construct the baseline model dataset.

[0083] The event identification and labeling module analyzes the changes in flow rate, valve opening, and motor speed over a continuous time period to determine whether the current operating state is in a quasi-steady state.

[0084] Within each monitoring window, the rate of change of flow rate, the rate of change of valve opening, and the rate of change of motor speed are calculated, and the magnitude of the changes in the rate of change of flow rate, the rate of change of valve opening, and the rate of change of motor speed are compared with preset stability thresholds.

[0085] When the rate of change of flow rate, the rate of change of valve opening, and the rate of change of motor speed are all within the allowable range and no rapid disturbance occurs, the current operating state is determined to be stable, and a quasi-steady-state dataset is output.

[0086] When the rate of change of flow rate, the rate of change of valve opening, and the rate of change of motor speed fluctuate briefly but quickly stabilize in the subsequent period, the corresponding period is identified as a quasi-steady state with acceptable disturbance.

[0087] When a sudden change in motor speed, a rapid adjustment in valve opening within a short period of time, a sudden fluctuation in differential pressure, or the triggering of a drive alarm is detected, the corresponding time segment is identified as an event segment and an event-labeled dataset is output. The event-labeled dataset contains information on the event type, duration, and direction of change of associated parameters.

[0088] The baseline energy efficiency modeling module and the event labeling module enable steady-state modeling and dynamic identification of motor operating states. The former automatically filters out steady-state operating ranges by detecting changes in flow rate, differential pressure, and rotational speed, extracts key energy efficiency parameters, and establishes energy efficiency curves, forming a baseline model dataset that reflects the relationship between energy consumption and water delivery, providing a standardized reference for subsequent energy efficiency assessments. The latter performs real-time analysis of the temporal changes in flow rate, valve opening, and rotational speed, distinguishing between steady-state, quasi-steady-state, and event stages, and outputs quasi-steady-state datasets and event-labeled datasets. This allows the system to automatically identify changes in operating states, filter out non-representative operating condition data, and ensure the stability and reliability of energy efficiency analysis results.

[0089] The event labeling module has an adaptive window update mechanism that monitors the change error between the flow dataset and the motor speed dataset in real time. When the change error between the two is detected to be continuously deviating from the threshold, the window is shortened; when the operating state is stable, the window is extended, and the output window is adjusted to record the dataset.

[0090] The energy efficiency estimation decomposition module performs consistency analysis on the correspondence between electrical energy input and hydraulic output in the quasi-steady-state interval, and extracts the active power and flow data at the corresponding time from the aligned dataset.

[0091] By statistically analyzing the changing trend of the ratio of active power to flow rate within a time window, a dataset of unit transmission energy consumption is output.

[0092] The overall efficiency is calculated based on the matching relationship between flow rate and pressure difference, and the energy efficiency assessment dataset is output.

[0093] Based on the multi-source operating characteristics, the physical sources of energy efficiency deviation are distinguished, and the harmonic current spectrum is combined with the changes in motor housing temperature and stator winding temperature to identify the changing trends of copper loss and iron loss caused by electromagnetic load.

[0094] The radial vibration acceleration and envelope spectrum characteristics are monitored and used to identify the trend of mechanical loss caused by mechanical friction, bearing wear or hydraulic pulsation. By comparing the change amplitude of different loss sources within the same analysis window, copper loss distribution dataset, iron loss distribution dataset and mechanical loss distribution dataset are output.

[0095] The copper loss distribution dataset, iron loss distribution dataset, and mechanical loss distribution dataset are merged and output as the loss decomposition result.

[0096] The current flow rate, differential pressure, and motor speed parameters are compared with the optimal efficiency point operating characteristics recorded in the baseline energy efficiency model. The deviation between the operating point and the optimal efficiency point is calculated, and the efficiency deviation dataset is output.

[0097] The energy efficiency estimation decomposition module dynamically adjusts the weights based on the changing trends of the temperature dataset, vibration feature dataset, and harmonic current spectrum dataset. It also allocates the weight ratios of copper loss, iron loss, and mechanical loss in real time according to the fluctuation amplitude and stability of various features within the same time window.

[0098] The energy efficiency estimation decomposition module enables accurate assessment of motor operating energy efficiency and structured decomposition of loss sources. Within the quasi-steady-state range, the module performs consistency analysis on the matching relationship between electrical energy input and hydraulic output, generating a unit transmission energy consumption dataset and an energy efficiency assessment dataset to quantify energy utilization levels under different operating conditions. Combining harmonic current spectra, temperature, and vibration characteristic signals, the module can distinguish between electromagnetic and mechanical loss sources, outputting copper loss, iron loss, and mechanical loss distribution datasets, and revealing the specific components of energy efficiency deviation through loss decomposition results. Simultaneously, the module compares real-time operating parameters with the optimal efficiency point of the baseline model, outputting an efficiency deviation dataset, achieving quantitative expression of energy efficiency change trends and dynamic monitoring of operating efficiency, providing a basis for energy efficiency optimization and maintenance decisions.

[0099] The source attribution assessment module identifies the specific sources of energy efficiency deviations in the loss decomposition results and calculates a credibility score based on multi-source evidence information.

[0100] Within the same analysis window, compare the copper loss distribution dataset, iron loss distribution dataset, and mechanical loss distribution dataset to calculate the degree of correspondence between each loss channel and the change in overall efficiency.

[0101] When the copper loss ratio increases significantly and is accompanied by an increase in the temperature of the motor housing or stator winding, the output deviation is attributed to an increase in electromagnetic load or winding aging.

[0102] When the iron loss ratio increases and the low-frequency distortion in the harmonic current spectrum is enhanced, the output deviation is attributed to changes in magnetic flux distribution or unstable drive control.

[0103] When the mechanical loss ratio increases and characteristic peaks appear in the radial vibration acceleration or envelope spectrum characteristics, the output deviation is attributed to bearing wear, rotor imbalance, or pump body hydraulic fluctuations.

[0104] The deviation attribution is correlated with the corresponding copper loss distribution dataset, iron loss distribution dataset, and mechanical loss distribution dataset to output an anomaly source determination dataset.

[0105] The loss decomposition results, temperature dataset, vibration feature dataset, and hydraulic parameter dataset are compared to analyze the consistency between information from different channels and quantify the reliability of the deviation analysis results.

[0106] When multiple data channels reflect the same energy efficiency deviation trend or point to the same source of deviation within the same time period, a high reliability score is output.

[0107] When there are differences between data from different channels or when the data is affected by noise, a low confidence score is output.

[0108] The scoring results are corrected by combining the running background in the event labeling dataset, distinguishing between transient disturbances and real anomalies, and outputting a credibility score dataset.

[0109] The source attribution assessment module performs a lag compensation process when generating credibility scores. It corrects the step size of the credibility score by analyzing the difference between the time of the event and the time of energy efficiency recovery.

[0110] The source attribution assessment module enables intelligent identification and reliable quantification of the causes of energy efficiency deviations. Based on loss decomposition results, the module comprehensively analyzes the changing characteristics of copper loss, iron loss, and mechanical loss, combining multi-source data such as temperature, vibration, and harmonic current spectrum to determine the specific source of energy efficiency decline. When aging of electromagnetic loads or windings leads to increased copper loss, magnetic flux instability causes increased iron loss, or bearing wear or hydraulic fluctuations exacerbate mechanical loss, the system can automatically output corresponding abnormal attribution results. Through comparative analysis of the consistency of multi-channel data, the module further generates a reliability score dataset to quantify the reliability of diagnostic results. Combining event labeling information and hysteresis compensation processes, the system can distinguish between transient disturbances and genuine anomalies, achieving accurate attribution and high-reliability output of energy efficiency deviations, significantly improving the diagnostic accuracy and stability of the monitoring system.

[0111] During continuous operation of the equipment, the online calibration and recharge module dynamically corrects the baseline model dataset based on the energy efficiency performance under quasi-steady-state operating conditions.

[0112] Active power, flow rate, differential pressure, and motor speed data are extracted from the aligned dataset and compared with the reference operating point in the baseline model dataset.

[0113] When the difference is within the normal fluctuation range, record the steady-state point distribution and output the steady-state verification dataset.

[0114] When the difference continues to shift but does not reach the abnormal limit, the energy efficiency reference value of the corresponding operating point is adjusted in small steps according to the direction and magnitude of the shift, and the updated baseline model dataset is output.

[0115] After equipment inspection, maintenance, or component replacement, representative operating points are extracted from the stable interval as strongly labeled data. The strongly labeled data is used as a high-confidence reference to correct the feature values ​​of the corresponding segments in the baseline model, and the corrected model dataset is output.

[0116] For each model revision, a version identifier is generated and the revision time, basis, and applicable scope are recorded, and the model version record dataset is output.

[0117] The online calibration and recharge module enables dynamic self-learning and continuous optimization of the motor energy efficiency baseline model. During normal equipment operation, the module corrects the baseline model dataset in real time based on the actual energy efficiency performance under quasi-steady-state conditions. By comparing the differences in active power, flow rate, pressure difference, and speed, it automatically determines the degree of deviation between the operating point and the baseline reference value. When the operating deviation is within the normal range, the system records the steady-state point distribution for model verification; when the deviation persists, the system adjusts the energy efficiency reference value in small steps according to the offset trend to keep the model consistent with the actual state of the equipment. After equipment maintenance or component replacement, the module updates the model features using high-confidence operating point data and generates model records with version identifiers, achieving traceable maintenance and long-term accuracy assurance of the energy efficiency model.

[0118] This invention simultaneously collects multi-source datasets and forms a unified aligned dataset through a time alignment mechanism, enabling consistent analysis of cross-domain signals. It automatically generates and updates baseline model datasets within steady-state and quasi-steady-state intervals. Through online calibration and model refeedback mechanisms, it continuously corrects the energy efficiency baseline, achieving dynamic consistency of the model as equipment ages, load changes, and maintenance adjustments. In the energy efficiency estimation decomposition module, it integrates electrical, vibration, and temperature characteristics to decompose energy consumption components, outputting copper loss, iron loss, and mechanical loss distribution datasets respectively. This allows for quantitative traceability of energy efficiency deviation sources. Through multi-channel consistency analysis in the source attribution assessment module, it cross-validates deviation results with temperature, vibration, and hydraulic data, automatically generating anomaly source determination and credibility score datasets, reducing false alarms and improving diagnostic reliability. Based on an adaptive window update mechanism, it can automatically adjust the data analysis time window according to changes in operating status, ensuring that energy efficiency analysis is only performed under stable operating conditions. Each model update generates a version record dataset, achieving traceability of model corrections and energy efficiency history, ensuring the consistency and credibility of monitoring results during long-term operation.

[0119] In summary, this system represents a technological leap from single-dimensional energy efficiency monitoring to multi-dimensional energy efficiency understanding. It can not only reflect the overall energy efficiency level of the motor in real time, but also locate the physical source of efficiency deviations and perform dynamic self-correction, significantly improving energy utilization efficiency and the level of intelligent equipment operation.

[0120] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0121] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A sensor-based motor energy efficiency real-time monitoring system, characterized in that, The system comprises a data collection alignment module, a baseline energy efficiency modeling module, an event identification labeling module, an energy efficiency estimation decomposition module, a source attribution evaluation module, and an online calibration backfilling module. The data collection alignment module collects electric parameter data sets, temperature data sets, mechanical operation data sets, hydraulic parameter data sets, heat exchange data sets, and control and environmental data sets, synchronizes them with the system control clock as the unified time reference, divides the sampling sequence into windows, and aligns the features, and outputs the aligned data sets. The baseline energy efficiency modeling module detects changes in flow, pressure difference, and rotating speed based on the aligned data sets, determines that it is in the steady-state operation interval when the change amplitude is within the stable threshold, extracts the corresponding data to establish the energy efficiency curve, and generates the baseline model data set. The event identification labeling module compares the flow, valve opening, and rotating speed change rate, outputs the quasi-steady-state data set when it is in a stable state, and outputs the event labeling data set when it detects start-stop, pressure difference mutation, or rapid valve adjustment. The energy efficiency estimation decomposition module calculates the comprehensive efficiency in the quasi-steady-state interval and outputs the energy efficiency evaluation data set, analyzes the energy efficiency deviation source based on temperature, vibration, and harmonic characteristics, and outputs the loss decomposition result and efficiency deviation degree data set. The source attribution evaluation module identifies the energy efficiency deviation source based on the loss decomposition result and generates the abnormal source judgment and credibility score data set. The online calibration backfilling module corrects the baseline model data set according to the quasi-steady-state operation performance. The energy efficiency estimation decomposition module analyzes the consistency between the corresponding relationship of electric energy input and hydraulic output in the quasi-steady-state interval, extracts the active power and flow data at the corresponding time from the aligned data set. The unit energy consumption data set is output by statistically analyzing the change trend of the active power to flow ratio in the time window. The comprehensive efficiency is calculated according to the matching relationship between flow and pressure difference, and the energy efficiency evaluation data set is output. The physical source of energy efficiency deviation is distinguished based on multi-source operation characteristics, the harmonic current spectrum is combined with the motor shell temperature and stator winding temperature change to identify the copper loss and iron loss change trend caused by electromagnetic load; The radial vibration acceleration and envelope spectrum characteristics are monitored, which are used to identify the mechanical friction, bearing wear, or hydraulic pulsation caused mechanical loss change trend, and the copper loss distribution data set, iron loss distribution data set, and mechanical loss distribution data set are output by comparing the change amplitudes of different loss sources in the same analysis window. The copper loss distribution data set, iron loss distribution data set, and mechanical loss distribution data set are combined and output as the loss decomposition result. The current flow, pressure difference, and motor rotating speed parameters are compared with the best efficiency point operation characteristics recorded in the baseline energy efficiency model, the deviation degree between the operating point and the best efficiency point is calculated, and the efficiency deviation degree data set is output. The source attribution evaluation module identifies the specific source of energy efficiency deviation in the loss decomposition result, and calculates the credibility score based on multi-source evidence information. The copper loss distribution data set, iron loss distribution data set, and mechanical loss distribution data set are compared in the same analysis window, and the corresponding degree between each loss channel and the comprehensive efficiency change is calculated. When the copper loss ratio rises significantly and the motor housing temperature or stator winding temperature increases, the output deviation is caused by electromagnetic load increase or winding aging; When the iron loss ratio rises and the low-frequency distortion in the harmonic current spectrum increases, the output deviation is caused by magnetic flux distribution change or unstable driving control; When the mechanical loss ratio rises and the radial vibration acceleration or characteristic peak appears in the envelope spectrum feature, the output deviation is caused by bearing wear, rotor imbalance or pump body hydraulic pressure fluctuation; Correlate the deviation attribution with the corresponding copper loss distribution dataset, iron loss distribution dataset and mechanical loss distribution dataset, and output the abnormal source judgment dataset; Compare the loss decomposition result, temperature dataset, vibration feature dataset and hydraulic parameter dataset, analyze the consistency degree between different channel information, and quantify the reliability of the deviation analysis result; When multiple data channels reflect the same energy efficiency deviation trend or point to the same deviation source at the same time period, output a high credibility score; When there are differences between different channel data or affected by noise, output a low credibility score; Combine the running background in the event labeling dataset to correct the scoring result, distinguish transient disturbance from real abnormality, and output the credibility score dataset.

2. A sensor-based motor efficiency real-time monitoring system as claimed in claim 1, wherein, The data collection alignment module collects three-phase voltage phasor, three-phase current phasor, power factor, active power, reactive power and harmonic current spectrum, and outputs electrical parameter dataset; Collect the motor housing temperature and stator winding temperature, and output temperature dataset; Collect the motor speed, and output mechanical operation dataset; Collect the flow, pressure difference and valve opening, and output hydraulic parameter dataset; Collect the water supply temperature and return water temperature, and output heat exchange dataset; Collect the internal torque estimation of the driver, the switching frequency of the driver, the alarm state of the driver and the environmental temperature, and output control and environmental dataset; Combine the electrical parameter dataset, temperature dataset, mechanical operation dataset, hydraulic parameter dataset, heat exchange dataset and control and environmental dataset to output the original measurement dataset; Obtain axial vibration acceleration and radial vibration acceleration, and identify characteristic energy modes of the axial vibration acceleration and radial vibration acceleration through envelope spectrum calculation method to form vibration feature dataset, wherein the characteristic energy modes include axial vibration acceleration and radial vibration acceleration corresponding to mechanical wear, bearing damage or assembly eccentricity; Take the system control clock as a unified time reference, divide the sampling sequences of the original measurement dataset and the vibration feature dataset into windows, establish a first correspondence relationship between the electrical parameter dataset, the vibration feature dataset and the hydraulic parameter dataset, and output the aligned dataset.

3. A sensor-based motor efficiency real-time monitoring system as claimed in claim 2, wherein, The baseline energy efficiency modeling module detects the variation amplitudes of the flow, pressure difference and motor speed in the preset time window in the aligned dataset, and when the variation amplitudes of the flow, pressure difference and motor speed are all in the set stable interval, determines that the corresponding period is a steady-state running interval; In the steady-state running interval, extract the active power, reactive power and hydraulic parameter dataset from the aligned dataset, record the correspondence relationship between the active power, reactive power and hydraulic parameter dataset, and output the steady-state running dataset; The steady-state operation data set is associated to output an energy efficiency curve data set, which reflects the relationship between water delivery and power consumption; The energy efficiency curve data set and the steady-state operation data set are jointly constructed into a baseline model data set.

4. A sensor-based motor efficiency real-time monitoring system as claimed in claim 3, wherein, The identification event labeling module analyzes the changes of flow, valve opening and motor speed in a continuous time period to determine whether the current operation state is in quasi-steady state: The flow rate of change, the valve opening rate of change and the motor speed rate of change are calculated in each monitoring window, and the change amplitudes of the flow rate of change, the valve opening rate of change and the motor speed rate of change are compared with the preset stability threshold; When the flow rate of change, the valve opening rate of change and the motor speed rate of change are all within the allowable range and no rapid disturbance occurs, it is determined that the current operation state tends to be stable, and a quasi-steady state data set is output; When the flow rate of change, the valve opening rate of change and the motor speed rate of change appear short-term fluctuations but quickly return to stability in the subsequent time period, the corresponding time period is identified as quasi-steady state with acceptable disturbance; When the motor speed jump, the valve opening adjustment in a short time, the pressure difference sudden fluctuation or the driver alarm state trigger are detected, the corresponding time segment is determined as an event segment and an event labeling data set is output, which includes event type, duration and associated parameter change direction information.

5. A sensor-based motor efficiency real-time monitoring system as claimed in claim 4, wherein, The online calibration recharge module dynamically corrects the baseline model data set according to the energy efficiency performance under quasi-steady state operation conditions during continuous operation of the equipment; The active power, flow, pressure difference and motor speed data are extracted from the aligned data set and compared with the reference operating point in the baseline model data set; When the difference is within the normal fluctuation range, the steady-state point distribution is recorded and a steady-state calibration data set is output; When the difference continuously deviates and does not reach the abnormal limit, the energy efficiency reference value of the corresponding operating point is adjusted in small steps according to the deviation direction and amplitude, and an updated baseline model data set is output; After equipment repair, maintenance or component replacement, representative operating points in the stable interval are extracted as strong label data, which are used as high confidence reference to correct the feature values of the corresponding section in the baseline model, and a corrected model data set is output; A version identifier is generated for each model correction, and the correction time, basis and applicable operating range are recorded, and a model version record data set is output.

6. A sensor-based motor efficiency real-time monitoring system as claimed in claim 5, wherein, The energy efficiency estimation decomposition module dynamically adjusts the weights based on the change trend of the temperature data set, the vibration feature data set and the harmonic current spectrum data set, and real-time allocates the weight proportion of copper loss, iron loss and mechanical loss according to the fluctuation amplitude and stability of each feature in the same time window.

7. A sensor-based motor efficiency real-time monitoring system as claimed in claim 6, wherein, The identification event labeling module has an adaptive window updating mechanism, which monitors the change error of the flow data set and the motor speed data set in real time, shortens the window when the change error of both continues to deviate from the threshold, and lengthens the window when the operation state is stable, and outputs a window adjustment record data set.

8. A sensor-based motor efficiency real-time monitoring system as claimed in claim 7, wherein, The source attribution evaluation module performs a lag compensation process when generating the credibility score, and corrects a change step of the credibility score by analyzing a difference between an event occurrence time and an energy efficiency recovery time.

Images