Energy-saving water pump parameter adaptive scheduling method and system

Abstract

This invention discloses an energy-saving water pump parameter adaptive scheduling method and system, relating to the field of water pump control technology. The method includes: collecting inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data; performing time alignment processing; extracting response features from the time-aligned water pump operating condition data to obtain dynamic response features; performing dynamic behavior distortion analysis based on a dynamic response reference model and the dynamic response features to obtain dynamic behavior distortion analysis results; if the dynamic behavior distortion analysis results indicate the existence of behavior distortion, then canceling the modification of water pump parameters by the multiplicative correction coefficient. This invention can combine water pump operating condition data and dynamic response features to analyze behavior distortion and cancel water pump parameter modifications, thereby achieving adaptive scheduling of water pump parameters, improving accuracy and stability, and reducing energy consumption per unit of water volume.

Description

Technical Field

[0001] This invention relates to the field of water pump control technology, and in particular to an energy-saving water pump parameter adaptive scheduling method and system. Background Technology

[0002] In the operation of secondary water supply pumping stations in residential communities, the outlet pressure and pipeline flow rate are usually used as control targets. Existing systems use frequency converters to adjust the pump speed to reduce energy consumption while meeting the terminal pressure requirements. However, in actual operation, due to the combined effects of various factors such as sensor installation deviations, valve mechanical characteristics, pump performance registration errors, filter weight assumptions, and heat dissipation protection, the system may misjudge the outlet pressure and pipeline flow rate, resulting in low accuracy, unstable water supply, and high energy consumption per unit volume of water.

[0003] In summary, the technical problems existing in the relevant technologies need to be improved. Summary of the Invention

[0004] The main objective of this invention is to propose an energy-saving water pump parameter adaptive scheduling method and system. This method can combine water pump operating condition data and dynamic response characteristics to analyze behavioral distortions and eliminate water pump parameter modifications, thereby achieving adaptive scheduling of water pump parameters, improving accuracy and stability, and reducing energy consumption per unit of water volume.

[0005] On one hand, embodiments of the present invention provide an adaptive scheduling method for energy-saving water pump parameters, comprising the following steps:

[0006] Collect inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data;

[0007] Time alignment processing is performed on the active power data of the frequency converter, the pipeline pressure data, the pipeline flow data, and the motor speed data;

[0008] Response features are extracted from the time-aligned water pump operating condition data to obtain dynamic response features. The dynamic response features include response delay, rise slope, curve integral, and response sequence. The water pump operating condition data includes the inverter active power data, the pipeline pressure data, the pipeline flow data, or the motor speed data.

[0009] Based on the dynamic response reference model and the dynamic response characteristics, dynamic behavior distortion analysis is performed to obtain the dynamic behavior distortion analysis results.

[0010] If the dynamic behavior distortion analysis results indicate the presence of behavior distortion, then the modification of the pump parameters by the multiplicative correction coefficient is cancelled.

[0011] On the other hand, embodiments of the present invention provide an energy-saving water pump parameter adaptive scheduling system, comprising:

[0012] The data acquisition module is used to collect inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data;

[0013] The time alignment module is used to perform time alignment processing on the inverter active power data, the pipeline pressure data, the pipeline flow data, and the motor speed data;

[0014] The response feature extraction module is used to extract response features from the time-aligned pump operating condition data to obtain dynamic response features. The dynamic response features include response delay, rise slope, curve integral, and response sequence. The pump operating condition data includes the inverter active power data, the pipeline pressure data, the pipeline flow data, or the motor speed data.

[0015] The behavior distortion analysis module is used to perform dynamic behavior distortion analysis based on the dynamic response reference model and the dynamic response characteristics, and obtain the dynamic behavior distortion analysis results.

[0016] The parameter control module is used to cancel the modification of the pump parameters by the multiplicative correction coefficient if the dynamic behavior distortion analysis result indicates the existence of behavior distortion.

[0017] The embodiments of this application include at least the following beneficial effects: First, the active power data of the frequency converter, pipeline pressure data, pipeline flow data, and motor speed data are collected and time-aligned. Then, the response features of the time-aligned pump operating condition data are extracted to obtain dynamic response features. Next, dynamic behavior distortion analysis is performed based on the dynamic response reference model and dynamic response features to obtain the dynamic behavior distortion analysis results. If the dynamic behavior distortion analysis results indicate the existence of behavior distortion, the modification of the pump parameters by the multiplicative correction coefficient is cancelled. This allows for the analysis of behavior distortion by combining the pump operating condition data and dynamic response features, and the cancellation of pump parameter modification, thereby achieving adaptive scheduling of pump parameters, improving accuracy and stability, and reducing energy consumption per unit of water volume.

[0018] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the description and the drawings. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below.

[0020] Figure 1 This is a flowchart of an energy-saving water pump parameter adaptive scheduling method according to an embodiment of the present invention;

[0021] Figure 2 This is a schematic diagram of the structure of an energy-saving water pump parameter adaptive scheduling system according to an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments.

[0023] In related technologies, during the operation of secondary water supply pumping stations in residential communities, the outlet pressure and pipeline flow rate are typically used as control targets. Existing systems use frequency converters to adjust pump speed to reduce energy consumption while meeting terminal pressure requirements. If a set of seemingly independent but causally related deviations occurs, conventional adjustment rules may fall into repeated adjustments, leading to increased energy consumption per unit volume of water. First, suppose that after routine maintenance, the straight pipe section upstream of a flow sensor is shortened, and a short-distance elbow is added to accommodate the pipeline layout. Under high flow velocity conditions, the rotating flow at the elbow causes a systematically higher instantaneous response from the sensor. This deviation is interpreted as the sensor's characteristics being normal, but the reading being a fixed proportion higher than the actual flow rate. In engineering, this situation can be caused by limited installation space, pipeline modification practices, or inconsistent understanding of straight pipe length standards by maintenance personnel. The physical mechanism of this change can be explained by the turbulent fluctuations in the flow field causing the static pressure distribution at the sensor's measuring orifice to deviate from the ideal distribution, thus resulting in a higher output value from the conversion formula based on pressure difference or velocity measurement, leading to low accuracy.

[0024] In the same control loop, if the control strategy is set to "pressure priority, flow assist," then when the flow reading is too high, the regulation logic will determine that the system supply is sufficient and attempt to reduce the flow rate to save energy. The logic here relies on the flow reading being used to confirm the system's margin in order to take aggressive energy-saving actions. The reduction in flow rate will cause a drop in pressure at the high-rise residential units; when insufficient pressure is detected, the controller immediately increases the speed to restore pressure, resulting in continuous fluctuations in the water supply within a short period, leading to unstable water supply and high energy consumption per unit of water.

[0025] In the above situation, the system needs to be able to identify and distinguish the systematically high flow reading caused by sensor installation during operation, and adjust the pump selection and speed setting online based on the actual efficiency of the pump set and the thermal margin of the motor during operation, so as to ensure that the terminal pressure meets the requirements and the energy consumption per unit water volume does not increase.

[0026] In view of this, this application acquires key information about pump operation through multi-source data acquisition, including inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data. After time alignment processing, these data form consistent pump operating condition data, laying the foundation for subsequent analysis. Next, by extracting response features from these operating condition data, dynamic response features including response delay, rise slope, curve integral, and response sequence are obtained. These features can meticulously characterize the behavior patterns of the pump system during dynamic changes.

[0027] Subsequently, this application compares these actually measured dynamic response characteristics with a pre-established dynamic response reference model to perform dynamic behavior distortion analysis. The dynamic response reference model represents the expected behavior of the pump system under ideal or normal operating conditions. By comparison, the deviation between the actual operating behavior and the ideal behavior can be identified, i.e., "behavioral distortion." This distortion may originate from various potential problems such as sensor drift, valve sticking, and pump performance degradation.

[0028] When dynamic behavior distortion analysis indicates the presence of behavior distortion, this application will cancel the modification of pump parameters by the multiplicative correction factor. In traditional energy-saving scheduling, the multiplicative correction factor is often used to fine-tune parameters such as pump speed based on real-time operating conditions to achieve energy saving. However, if the system itself has behavior distortion, these corrections based on erroneous or inaccurate information will not only fail to achieve energy saving but may also further deteriorate water supply stability or energy consumption performance. For example, if the flow sensor reading is systematically high, the control system may misjudge that the water supply is abundant and reduce the speed. If the multiplicative correction factor further reduces the speed, it will cause the pressure in high-rise residents to drop, triggering a vicious cycle of repeated speed increases and decreases. By canceling the multiplicative correction factor, this application can avoid further negative impacts on the system caused by erroneous parameter adjustment commands, buying time for subsequent fault diagnosis and problem solving.

[0029] The embodiments of this application will be explained in detail below with reference to the accompanying drawings:

[0030] Figure 1 This is an optional flowchart of an energy-saving water pump parameter adaptive scheduling method provided in an embodiment of this application. Figure 1 The method may include, but is not limited to, steps S101 to S105.

[0031] Step S101: Collect inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data;

[0032] Step S102: Perform time alignment processing on the inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data;

[0033] Step S103: Extract response features from the time-aligned pump operating condition data to obtain dynamic response features. The dynamic response features include response delay, rise slope, curve integral, and response sequence. The pump operating condition data includes inverter active power data, pipeline pressure data, pipeline flow data, or motor speed data.

[0034] Step S104: Based on the dynamic response reference model and dynamic response characteristics, perform dynamic behavior distortion analysis to obtain the dynamic behavior distortion analysis results;

[0035] Step S105: If the dynamic behavior distortion analysis result indicates the existence of behavior distortion, then cancel the modification of the pump parameters by the multiplicative correction coefficient.

[0036] Steps S101 to S105 as shown in the embodiments of this application can combine water pump operating condition data and dynamic response characteristics to analyze behavioral distortion and cancel water pump parameter modification, so as to realize adaptive scheduling of water pump parameters, improve accuracy and stability, and reduce energy consumption per unit water volume.

[0037] In some embodiments, steps S101-S105 may involve collecting inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data. These data form the basis for evaluating the pump's operating status. For example, this can be achieved by installing sensors at key locations in the pumping station. Inverter active power data can be obtained through the inverter's built-in power measurement module; pipeline pressure data can be obtained through pressure sensors installed on the water supply pipeline, for example, near the pump outlet and remote user access points; pipeline flow data can be obtained through flow sensors installed on the main water supply pipeline, such as ultrasonic or electromagnetic flow meters; and motor speed data can be obtained through the inverter's internal speed feedback signal. These sensors can be standalone physical devices or virtual sensors integrated into existing control systems. It is understood that inverter active power data refers to the actual effective power output by the inverter during operation, reflecting the electrical energy consumed by the pump system. Pipeline pressure data refers to the fluid pressure inside the water supply pipeline, typically measured by pressure sensors, and is an important indicator of the water supply system's stability. Pipeline flow rate data refers to the volume of fluid flowing through a pipeline per unit time, usually measured by a flow sensor, and reflects the supply capacity of the water supply system. Motor speed data refers to the rotational speed of the pump drive motor, which is adjusted by a frequency converter and directly affects the pump's head and flow rate.

[0038] Then, time alignment is performed on the inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data. Since different sensors may have different sampling frequencies or data transmission delays, time alignment is a crucial step to ensure data consistency. For example, linear interpolation or nearest neighbor interpolation can be used to unify all data to the same timestamp. Specifically, an average sampling frequency can be calculated, and all data streams can be aligned to this average sampling frequency through interpolation or resampling. Alternatively, a precise timestamp can be recorded for each data point, and data from different data sources can be correlated through timestamp matching.

[0039] The time-aligned pump operating data is then processed to extract response features, yielding dynamic response features. These dynamic response features include response delay, rise slope, curve integral, and response sequence. The pump operating data includes inverter active power data, pipeline pressure data, pipeline flow data, or motor speed data. For example, when the pump speed changes, the curves monitoring the changes in pipeline pressure, flow rate, and active power are observed. Response delay can be determined by detecting the time difference between a change in the input signal and the first significant change in the output signal. The rise slope can be obtained by calculating the slope of the rising segment of the change curve. The curve integral can be calculated by integrating the change curve over a specific time period. The response sequence can be determined by comparing the time points at which different parameters reach their stable or peak values. It can be understood that response delay refers to the time interval between receiving a command and starting to respond. The rise slope refers to the rate at which the system parameter's value increases during the response process. The curve integral refers to the cumulative change of the system parameter over a certain time period. The response sequence refers to the order in which different parameters change during the response process.

[0040] Based on the dynamic response reference model and dynamic response characteristics, dynamic behavioral distortion analysis is performed to obtain the results. The dynamic response reference model can be a pre-established ideal pump operation model that describes the expected response of the pump system to changes in speed under normal operating conditions. For example, a dynamic response reference model can be obtained by mathematically modeling based on the pump's performance curve and pipeline characteristics. During dynamic behavioral distortion analysis, the actual extracted dynamic response characteristics can be compared with the dynamic response reference model. For example, the deviation between the actual response characteristics and the model's predicted characteristics can be calculated, or statistical methods (such as mean squared error and correlation coefficient) can be used to quantify the degree of distortion. If the deviation exceeds a preset threshold, behavioral distortion is considered to exist.

[0041] If the dynamic behavior distortion analysis indicates the presence of behavior distortion, then the modification of pump parameters using multiplicative correction factors should be cancelled. Multiplicative correction factors are typically used in energy-saving scheduling to fine-tune pump parameters to optimize energy consumption. However, when system behavior distortion exists, these correction factors may exacerbate rather than solve the problem. For example, if the system experiences misjudgments due to excessively high flow sensor readings, multiplicative correction factors may further reduce the pump speed, leading to insufficient water supply pressure. Therefore, in such cases, cancelling the application of multiplicative correction factors can prevent erroneous parameter adjustments from further deteriorating system performance. For example, the multiplicative correction factors can be reset to 1, or their effect can be temporarily disabled until the behavior distortion problem is resolved.

[0042] Through the above technical solution, this embodiment can capture subtle behavioral patterns of the water pump system during dynamic changes by collecting multi-dimensional operational data in real time and extracting dynamic response characteristics. By comparing these actual behaviors with the dynamic response reference model, it can accurately identify whether the system exhibits "behavioral distortion." This distortion analysis can not only detect obvious faults but also identify performance deviations caused by hidden problems such as sensor drift and slight valve sticking. Once behavioral distortion is identified, this embodiment can decisively cancel the modification of water pump parameters by the multiplicative correction coefficient. This strategy avoids parameter optimization based on potentially erroneous information when the system behavior is abnormal, thereby preventing erroneous correction instructions from further deteriorating water supply stability and energy consumption performance. For example, if the flow sensor reading is too high, causing the system to misjudge an abundant water supply, the traditional method might continue to reduce the speed through the multiplicative correction coefficient to "save energy," but this actually leads to a drop in pressure for residents in high-rise buildings. This embodiment can identify this behavioral distortion and cancel the speed reduction correction, thereby avoiding periodic pressure drops and compensation. Therefore, this embodiment demonstrates significant progress in improving the robustness and adaptability of water supply system operation.

[0043] In some embodiments, step S103 involves extracting response features from the time-aligned pump operating condition data to obtain dynamic response features, which may include, but is not limited to, the following steps:

[0044] Multi-level sliding median filtering is applied to the time-aligned pump operating data to remove instantaneous spikes and outliers.

[0045] Adaptive threshold processing is applied to the pump operation data after multi-layer sliding median filtering to identify normal and abnormal data segments.

[0046] Calculate the response delay, rise slope, and curve integral based on the normal data segment;

[0047] Identify adjacent data segments that are adjacent to the abnormal data segments;

[0048] Based on adjacent data segments, the abnormal data segments are smoothed by interpolation to obtain a smooth response curve;

[0049] Extract the response order from the smooth response curve.

[0050] In some embodiments, the time-aligned pump operating data can first undergo multi-level sliding median filtering to remove instantaneous spikes and outliers. The purpose is to effectively suppress instantaneous spikes and random noise that may occur during data acquisition, as well as outliers caused by sensor malfunctions or communication interference. For example, by applying median filtering at different time scales, the main trend of the data can be preserved while smoothing short-term fluctuations, thereby providing a cleaner and more reliable data foundation for subsequent feature extraction.

[0051] Then, adaptive thresholding is applied to the pump operating data after multi-layer sliding median filtering to identify normal and abnormal data segments. The aim is to accurately distinguish between normal operating conditions and potential abnormal fluctuations in the data. For example, the threshold can be dynamically adjusted based on the statistical characteristics of the data (e.g., mean, standard deviation, or rate of change) to identify data segments that deviate from the normal range and mark them as abnormal data segments, while the remaining portions are identified as normal data segments.

[0052] Based on the normal data segment, calculate the response delay, rise slope, and curve integral. Response delay can be understood as the time required for the system to respond to a change in input; it can be determined by detecting the time difference between the change in the input signal and the first significant change in the output signal. The rise slope refers to the rate of change of the system response from a certain reference value to another target value; it can be obtained by calculating the slope of the rising segment of the curve. The curve integral reflects the accumulated energy or effect during the response process; it can be calculated by integrating the curve over a specific time period.

[0053] Next, adjacent data segments are identified. The purpose is to provide contextual information for repairing the anomalous data segment. Adjacent data segments refer to the immediately preceding and following segments that are identified as normal. These segments reflect the normal operating trend of the system before and after the anomaly. Based on these adjacent data segments, smooth interpolation is performed on the anomalous data segment to obtain a smooth response curve. The aim is to utilize information from adjacent normal data segments to reasonably fill and correct the anomalous data segment, thereby eliminating the impact of outliers on the overall response curve shape. For example, linear interpolation, spline interpolation, or model-based prediction interpolation methods can be used to ensure that the interpolated data remains smooth and continuous with the preceding and following normal data segments.

[0054] Finally, the response sequence is extracted from the smoothed response curve. The response sequence refers to the order in which different operating condition data (such as pressure, flow rate, speed, and power) respond when the pump's operating state changes. By analyzing the order in which each parameter in the smoothed response curve reaches its stable value or a specific threshold, the response sequence can be accurately identified, which is crucial for understanding the dynamic characteristics of the pump system.

[0055] Through the above technical solutions, this embodiment effectively addresses common issues such as noise, spikes, and outliers in pump operating data, significantly improving the robustness and accuracy of dynamic response feature extraction. Specifically, the combination of multi-layer sliding median filtering and adaptive thresholding ensures data quality and avoids feature calculation biases caused by data contamination. Smoothing interpolation effectively repairs abnormal data segments, enabling the acquisition of continuous and reliable response curves even in cases of incomplete or anomaly-prone data, thus accurately extracting the response sequence. Therefore, this embodiment provides a more accurate and comprehensive characterization of pump dynamic behavior, offering a solid data foundation for adaptive scheduling of energy-saving pump parameters and facilitating more accurate identification of pump operating states and potential performance problems.

[0056] In some embodiments, in step S104, dynamic behavior distortion analysis is performed based on the dynamic response reference model and dynamic response characteristics to obtain the dynamic behavior distortion analysis results, which may include, but is not limited to, the following steps:

[0057] Step S201: Identify the pump's operating status based on the pump's operating condition data;

[0058] Step S202: If the water pump is in a stable operating state, calculate the water pump efficiency curve based on the water pump operating condition data.

[0059] Step S203: Based on the pump efficiency curve, perform local adjustments to the dynamic response reference model;

[0060] Step S204: During low load periods, the pump speed is perturbed to obtain micro-perturbation response characteristics, which include pressure response, power response and flow response.

[0061] Step S205: Based on the perturbation response characteristics, verify the effectiveness of the dynamic response reference model after local adjustment processing, and obtain the effectiveness verification results;

[0062] Step S206: Based on the validity verification results, dynamic response characteristics, and the dynamic response reference model after local adjustment, perform dynamic behavior distortion analysis to obtain the dynamic behavior distortion analysis results.

[0063] In some embodiments, the operating conditions and performance of the pump system may change over time, due to factors such as equipment wear, pipe scaling, or fluctuations in system demand. If the dynamic response reference model remains fixed, it may not accurately reflect the true dynamic behavior of the pump under the current operating conditions, leading to a decrease in the accuracy of the dynamic behavior distortion analysis results, and even potential misjudgments, thus affecting the reliability of the energy-saving pump parameter adaptive scheduling method.

[0064] Therefore, the pump's operating status can be identified first based on its operating data. Analyzing this data can determine whether the pump is currently operating stably or unstablely. This is because the dynamic response characteristics of a pump can differ significantly under different operating conditions. For example, its dynamic response differs from that during stable operation during startup, shutdown, or significant load changes, requiring differentiated handling. If the pump's operating status is stable, the pump efficiency curve can be calculated based on the operating data. The pump efficiency curve reflects the pump's energy conversion efficiency under different operating conditions, and its changes can indicate performance degradation or changes in system resistance, both of which affect the pump's dynamic response characteristics. For example, the ratio between pipeline flow and inverter active power can be calculated as the efficiency at the corresponding time point, and the efficiency at each time point can be combined to form the pump efficiency curve.

[0065] Then, based on the pump efficiency curve, the dynamic response reference model is locally adjusted. The purpose is to make the reference model better adapt to the current actual operating performance of the pump. For example, when the pump efficiency decreases, the parameters related to energy conversion efficiency in the reference model (such as the response slope and the tolerance range of the curve integral) will be adjusted accordingly to reflect the impact of efficiency changes on the dynamic response.

[0066] Then, during low-load periods, the pump speed is perturbed to obtain micro-perturbation response characteristics. These characteristics include pressure response, power response, and flow response, providing instantaneous feedback information on the pump's response to speed changes. Low-load periods typically indicate lower instantaneous demand on pump performance. At this time, small-amplitude speed perturbations can safely probe the pump's true dynamic response under current operating conditions without significantly impacting system operation. Based on the micro-perturbation response characteristics, the effectiveness of the locally adjusted dynamic response reference model is verified, obtaining validity verification results. This aims to confirm whether the reference model, adjusted by the efficiency curve, accurately reflects the pump's current dynamic behavior. By comparing the actually measured micro-perturbation response with the adjusted model's predicted response, the model's accuracy and applicability can be evaluated.

[0067] Finally, based on the validity verification results, dynamic response characteristics, and the dynamically adjusted reference model, dynamic behavior distortion analysis was performed to obtain the results. This means that the final distortion analysis not only considered the deviation between the pump's dynamic response characteristics and the reference model, but also incorporated the model's own validity assessment, thereby improving the accuracy and reliability of distortion judgment.

[0068] To illustrate this technical solution more clearly, a specific example is used below. Suppose that after long-term operation, the overall efficiency of a water pump in a water supply system decreases due to impeller wear. If the dynamic response reference model is based on the initial performance of the pump, then when the actual efficiency of the pump decreases, its dynamic response to control commands (such as changes in pressure response, flow response, and power response) may deviate from the predictions of the initial model. This deviation may be misjudged as "behavioral distortion," leading to the unnecessary cancellation of multiplicative correction factors to modify the pump parameters, thus affecting energy-saving scheduling. This embodiment first identifies that the pump is in a stable operating state based on the pump's operating condition data. Subsequently, based on the current operating condition data, the actual efficiency curve of the pump is calculated, and it is found that it has decreased compared to the initial efficiency curve. Based on this new efficiency curve, the parameters related to energy conversion efficiency in the dynamic response reference model (e.g., the expected range of the response slope, the tolerance of the curve integral) are locally adjusted to reflect the pump's current lower efficiency level.

[0069] During low-load periods, the system applies small-amplitude speed disturbances to the water pump and collects actual perturbation response characteristics (pressure response, power response, and flow response). These perturbation response characteristics are then used to validate the effectiveness of the locally adjusted dynamic response reference model. If the validation results show that the adjusted model can accurately predict the perturbation response of the water pump at the current efficiency, the model adjustment is considered effective. Finally, this locally adjusted and validated dynamic response reference model is used for dynamic behavioral distortion analysis. At this point, if the dynamic response characteristics of the water pump still deviate significantly from this updated model, it will be considered as exhibiting behavioral distortion. For example, if a pipeline leak causes an abnormally large increase in flow response, which contradicts the adjusted model's prediction, it will be accurately identified as behavioral distortion. Conversely, if the response change is merely due to efficiency degradation but conforms to the adjusted model, it will not be misjudged as distortion. In this way, this embodiment can distinguish between response changes caused by normal performance degradation and behavioral distortion caused by actual faults or anomalies, thereby improving the accuracy of the judgment.

[0070] Through the above technical solution, this embodiment can significantly improve the accuracy and reliability of dynamic behavior distortion analysis in the adaptive scheduling method for energy-saving water pump parameters. Since the dynamic response reference model can adaptively adjust according to the actual operating state and efficiency changes of the water pump, it can more accurately reflect the dynamic characteristics of the water pump under different operating conditions. This effectively avoids misjudgments that may be caused by a fixed model and reduces the risk of false distortion alarms or omissions of true distortions due to discrepancies between the model and reality. Furthermore, by introducing speed disturbances and model validity verification, the real-time monitoring and evaluation capability of the water pump system's dynamic behavior is further enhanced, enabling the system to identify potential pump failures or performance degradation more promptly and accurately. This provides a more reliable basis for canceling the multiplicative correction coefficient, ultimately improving the robustness and energy-saving effect of the entire energy-saving scheduling system.

[0071] In some embodiments, in step S201, identifying the pump operating status based on the pump operating condition data may include, but is not limited to, the following steps:

[0072] Step S301: Based on the pump operating condition data, calculate the average data fluctuation amplitude, average value, and standard deviation within the sliding time window;

[0073] Step S302: If the current data fluctuation amplitude is greater than the average data fluctuation amplitude, then the water pump operating status is determined to be unstable.

[0074] Step S303: If the current data fluctuation range is less than or equal to the average data fluctuation range, then the average value and standard deviation are judged.

[0075] Step S304: If the average value is less than the preset average threshold and the standard deviation is less than the preset deviation threshold, then monitor the grid voltage data and the upstream main pipe pressure data.

[0076] Step S305: Identify the operating status of the water pump based on the grid voltage data and the upstream main pipe pressure data.

[0077] In some embodiments, since state identification is based solely on the internal operating data of the pump, it may not fully reflect the actual operating environment of the pump. In particular, when there are external power grid fluctuations or upstream pressure disturbances, the internal data of the pump may appear to be stable, but in reality, it is not truly stable. This may lead to deviations in the adjustment and validity verification of the dynamic response reference model.

[0078] Therefore, based on the pump's operating data, the average data fluctuation amplitude, mean, and standard deviation within the sliding time window can be calculated. By calculating the average data fluctuation amplitude, mean, and standard deviation of these data within the sliding time window, the stability and dispersion of the pump's operation can be quantified. The length of the sliding time window can be set according to the actual application scenario and data sampling frequency to capture data characteristics at different time scales. The average data fluctuation amplitude is used to measure the overall drastic change in the data over a period of time.

[0079] If the current data fluctuation is greater than the average data fluctuation, it indicates significant fluctuations in the pump's operating status, thus confirming that the pump's operation is unstable. This rapid judgment mechanism can respond promptly to sudden changes in operating conditions. If the current data fluctuation is less than or equal to the average data fluctuation, a more refined judgment of the mean and standard deviation is required. The mean reflects the central trend of the data within the sliding time window, while the standard deviation reflects the dispersion of the data.

[0080] If the average value is less than the preset average threshold and the standard deviation is less than the preset deviation threshold, it indicates that the internal operating parameters of the water pump are relatively stable with small fluctuations. However, this does not immediately confirm a completely stable state, as external factors may cause a "pseudo-stability" phenomenon. Further monitoring of grid voltage data and upstream mains pressure data is possible. Grid voltage data reflects the stability of the power supply system, while upstream mains pressure data reflects the external fluid environment on the pump's suction side. More importantly, when the pump is in a stable operating state, a large amount of pump operating condition data can be collected as experimental data. Statistical analysis of the experimental data can be performed to calculate the average value as the preset average threshold and the standard deviation as the preset deviation threshold.

[0081] Finally, the pump's operating status is identified based on the grid voltage data and upstream mains pressure data. By comprehensively analyzing this external data, the pump's true operating status can be assessed more comprehensively, avoiding misjudging a "pseudo-stable" state affected by external disturbances as a true stable state.

[0082] To illustrate this technical solution more clearly, a specific example is used below. Suppose that for a certain period of time, the average fluctuation amplitude, average value, and standard deviation of the inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data calculated within a sliding time window are all below a preset threshold, indicating that the pump's internal operating parameters exhibit a relatively stable trend. Based solely on this internal data, the system might classify it as a stable state. However, this embodiment further monitors the grid voltage data and upstream mains pressure data. For example, if a slight but persistent harmonic distortion is detected in the grid voltage data, or periodic small pulsations are detected in the upstream mains pressure data, this embodiment can identify such potential external influences even if these external disturbances have not directly caused significant fluctuations in the pump's internal operating data. In this case, the system will not simply classify the pump as completely stable, but may instead identify it as a "pseudo-stable" state. This identification result will prompt the system to take a more cautious approach to adjusting and validating the dynamic response reference model in subsequent dynamic behavior distortion analysis. For example, it may not immediately carry out aggressive parameter optimization, but wait until the external environment stabilizes before doing so, thereby avoiding potential risks or inaccurate scheduling caused by misjudging the stable state.

[0083] By introducing the above technical solution, this embodiment effectively avoids misjudging a "pseudo-stable" state affected by external disturbances as a true stable state by introducing monitoring of grid voltage data and upstream main pipe pressure data. This accurate state identification capability can significantly improve the accuracy of subsequent local adjustments to the dynamic response reference model and the reliability of effectiveness verification, thereby ensuring that the energy-saving water pump parameter adaptive scheduling system can make more reasonable and effective parameter modification decisions under various complex operating conditions, ultimately improving the overall operating efficiency and stability of the water pump system.

[0084] In some embodiments, in step S305, identifying the pump operating status based on the grid voltage data and the upstream main pipe pressure data may include, but is not limited to, the following steps:

[0085] Harmonic analysis is performed on the grid voltage data to extract the harmonic distortion rate and instantaneous voltage drop amplitude.

[0086] Frequency analysis was performed on the upstream main pipe pressure data to identify periodic pulsations and sudden changes;

[0087] Calculate the instantaneous rate of change and trend of change based on the operating data of the water pump;

[0088] Anomaly analysis was performed on the grid voltage data based on harmonic distortion rate, instantaneous drop amplitude, and instantaneous change rate, resulting in the first anomaly analysis result.

[0089] Based on periodic pulsations, sudden changes, and trends, anomaly analysis was performed on the upstream main pipe pressure data to obtain the second anomaly analysis results.

[0090] If both the first and second anomaly analysis results show no anomalies, the pump's operating state is determined to be stable; otherwise, the pump's operating state is determined to be pseudo-stable.

[0091] In some embodiments, simple monitoring alone may not be able to fully reveal potential disturbances in the power grid or pipeline system, such as harmonic distortion of the grid voltage or periodic pulsations in the upstream mains pressure. These potential disturbances may cause the pump to be in a "pseudo-stable" state that appears stable but is actually affected by external factors. If this is misjudged as a true stable state, it may affect the accuracy of the subsequent dynamic response reference model and the effectiveness of adaptive scheduling of energy-saving pump parameters.

[0092] Therefore, harmonic analysis can be performed on the grid voltage data to extract the harmonic distortion rate and instantaneous voltage drop amplitude. Signal processing methods such as Fourier transform can be used to decompose the acquired grid voltage data into fundamental and harmonic components, from which the harmonic distortion rate and instantaneous voltage drop amplitude can be calculated. The harmonic distortion rate reflects the degree to which the grid voltage waveform deviates from a standard sine wave, while the instantaneous voltage drop amplitude indicates the degree of sudden voltage decrease within a short period. These parameters are key indicators for evaluating the power supply quality of the grid.

[0093] Then, frequency analysis is performed on the upstream main pipe pressure data to identify periodic pulsations and sudden changes. Fast Fourier Transform (FFT) or other spectral analysis techniques can be used to analyze the frequency components of the upstream main pipe pressure data to identify whether periodic pressure pulsations or sudden pressure changes exist. Periodic pulsations may originate from the operation of upstream pumping stations or valve vibrations, while sudden changes may indicate transient events in the piping system, such as rapid valve opening and closing or water hammer effects.

[0094] Then, based on the pump operating data, the instantaneous rate of change and trend are calculated. Differential calculations or moving averages can be performed on the inverter's active power data, pipeline pressure data, pipeline flow data, or motor speed data to obtain the rate and direction of change of these parameters over a short period. The instantaneous rate of change reflects the degree of fluctuation in the data at the current moment, while the trend reveals the overall direction of the data over a period of time.

[0095] Anomaly analysis is performed on the power grid voltage data based on harmonic distortion rate, instantaneous voltage drop amplitude, and instantaneous rate of change to obtain the first anomaly analysis result. The extracted harmonic distortion rate, instantaneous voltage drop amplitude, and instantaneous rate of change of the pump operating condition data can be compared with a first historical baseline. If any indicator exceeds the normal range, the power grid voltage data is determined to be abnormal. Historical power grid voltage data and pump operating condition data under abnormal conditions can be analyzed to calculate the harmonic distortion rate, instantaneous voltage drop amplitude, and instantaneous rate of change as the first historical baseline.

[0096] Anomaly analysis is performed on the upstream main pipe pressure data based on periodic pulsations, sudden changes, and trends to obtain a second anomaly analysis result. The identified periodic pulsations, sudden changes, and pump operating condition data trends can be compared with a second historical baseline. If significant periodic pulsations or sudden changes, or abnormal trends, are found, the upstream main pipe pressure data is determined to be abnormal. Historical upstream main pipe pressure data and pump operating condition data under abnormal conditions can be analyzed to calculate periodic pulsations, sudden changes, and trends as a second historical baseline.

[0097] If both the first and second anomaly analysis results show no anomalies, the pump's operating state is determined to be stable. Otherwise, the pump's operating state is determined to be pseudo-stable. By comprehensively judging the first and second anomaly analysis results, the pump's operating state can be identified more accurately. If both show no anomalies, the pump is considered to be in a truly stable state; conversely, if either analysis result shows an anomaly, even if the pump's own operating data appears stable, it is judged to be in a pseudo-stable state to avoid external factors misleading the pump's performance evaluation and scheduling strategy.

[0098] Through the above technical solution, this embodiment can more accurately and robustly identify the true operating state of the water pump, especially when the pump's own operating parameters fluctuate little. It can effectively distinguish between a true stable state and a pseudo-stable state affected by external disturbances from the power grid or pipeline system. This refined state identification avoids misjudging pseudo-stable states as stable states, thereby preventing local adjustments and validity verification of the dynamic response reference model based on inaccurate state information and improving the accuracy of dynamic behavior distortion analysis. Therefore, it ensures the effectiveness and reliability of subsequent energy-saving water pump parameter adaptive scheduling strategies, avoids scheduling deviations or efficiency losses caused by external hidden problems, and further improves the system's energy-saving effect and operational stability.

[0099] In some embodiments, in step S202, calculating the pump efficiency curve based on the pump operating condition data may include, but is not limited to, the following steps:

[0100] The pump operating condition data is processed by time series segmentation to obtain complete data segments, insufficient data segments, and unevenly distributed data segments;

[0101] The first efficiency curve is obtained by fitting the complete data segment.

[0102] Based on historical operating data, data supplementation is performed on segments with insufficient data volume;

[0103] The second efficiency curve is obtained by fitting the insufficient data segment after data supplementation.

[0104] The third efficiency curve is obtained by fitting the unevenly distributed data segment using the locally weighted linear regression method.

[0105] The first, second, and third efficiency curves are smoothed by interpolation to generate the pump efficiency curve.

[0106] In some embodiments, the collected pump operating condition data may have quality issues such as missing data, insufficient data volume, or uneven data distribution. Directly fitting these raw data with quality problems may result in inaccurate or unrepresentative pump efficiency curves, which in turn affects the accuracy of subsequent local adjustments to the dynamic response reference model. Ultimately, this may lead to deviations in the dynamic behavior distortion analysis results, thereby affecting the effectiveness of adaptive scheduling of energy-saving pump parameters.

[0107] Therefore, the pump operating condition data can be first processed by time series segmentation to obtain complete data segments, insufficient data segments, and unevenly distributed data segments. Time series segmentation aims to divide the original continuous pump operating condition data into different segments based on the quality and characteristics of the data. Among them, the complete data segment refers to an ideal data segment with continuous, unmissing, and evenly distributed data points; the insufficient data segment refers to a data segment with sparse data points or insufficient sampling within a specific operating range; and the unevenly distributed data segment refers to a data segment where data points exist but are irregularly distributed, possibly exhibiting localized dense or sparse phenomena.

[0108] Then, the complete data segment is fitted to obtain the first efficiency curve. For the complete data segment, due to its high data quality, conventional curve fitting methods, such as polynomial fitting or spline fitting, can be directly used to obtain the first efficiency curve. The purpose is to accurately capture the efficiency characteristics of the water pump under stable and fully sampled operating conditions.

[0109] Then, based on historical operating data, data supplementation is performed on segments with insufficient data. To compensate for the fitting difficulties caused by data sparsity, historical operating data can be used to supplement the data for segments with insufficient data. This historical operating data can be valid data selected from long-term operating data accumulated by the same or similar pumps under similar operating conditions. The purpose of data supplementation is to increase the density of data points, so that the subsequent fitting process can more accurately reflect the efficiency characteristics of the pump, and the supplemented data segment with insufficient data is fitted to obtain the second efficiency curve.

[0110] A third efficiency curve is obtained by fitting the data to the unevenly distributed sub-segment using locally weighted linear regression. For unevenly distributed sub-segments, traditional global fitting methods may fail to accurately capture local characteristics due to the irregular distribution of data points. Therefore, locally weighted linear regression is employed. By assigning different weights to each data point, data points closer to the current fitting point have a greater influence, thus better adapting to local data variations and yielding the third efficiency curve.

[0111] Finally, the first, second, and third efficiency curves are smoothed and interpolated to generate the pump efficiency curve. The purpose of the smoothing interpolation is to eliminate discontinuities or abrupt changes between the fitting results of different segments, and to generate a continuous, smooth pump efficiency curve that can comprehensively reflect the efficiency characteristics of the pump under various operating conditions.

[0112] To illustrate this technical solution more clearly, a specific example is used below. Suppose that in an industrial circulating water system, long-term operational data collection is performed on an energy-saving water pump. In the operational data of a certain month, after time alignment processing of the inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data, it is found that: some time periods (e.g., continuous operation for 10 days, 24 hours a day) have stable, complete, and evenly distributed data collection, identified as complete data segments. For this part of the data, the least squares method can be used for polynomial fitting to obtain the first efficiency curve, accurately reflecting the efficiency characteristics of the water pump under stable operating conditions. Another part of the data period (e.g., during low-load operation at night, when the data collection frequency decreases or some sensors intermittently malfunction) has significantly insufficient data, identified as insufficient data segments. For this part of the data, the system automatically retrieves historical operational data of the water pump under similar low-load conditions over the past year, filters and supplements it to increase data density. Subsequently, spline fitting is performed on the supplemented data to obtain the second efficiency curve. In some time periods (e.g., when the system frequently starts and stops or experiences severe load fluctuations), although data exists, its distribution is extremely uneven. Data is very dense at some speeds or flow rates, while very sparse at others, and these are identified as unevenly distributed sub-segments. For this portion of data, a locally weighted linear regression method is used for fitting, allowing the fitted curve to better adapt to the local characteristics of the data points, resulting in a third efficiency curve. Finally, the efficiency curves obtained by these three different methods are smoothed and interpolated, for example using cubic spline interpolation, to eliminate the lack of smoothness between different curves, ultimately generating a continuous and smooth pump efficiency curve covering the entire operating range of the pump. This comprehensive efficiency curve will be used for subsequent local adjustments to the dynamic response reference model, ensuring accurate dynamic behavior distortion analysis under various complex operating conditions.

[0113] Through the above technical solution, this embodiment effectively addresses the complex and ever-changing data quality issues in actual operation by employing intelligent time series segmentation and targeted fitting strategies, ensuring the generation of a high-confidence pump efficiency curve under various operating conditions. This accurate and robust efficiency curve more realistically reflects the actual operating performance of the pump, thus providing a more accurate basis for local adjustments to the dynamic response reference model. This, in turn, improves the accuracy and reliability of the entire energy-saving pump parameter adaptive scheduling method, ultimately achieving optimized energy-saving effects and more stable system operation.

[0114] In some embodiments, in step S203, the dynamic response reference model is locally adjusted based on the pump efficiency curve, which may include, but is not limited to, the following steps:

[0115] The pump efficiency curve is segmented by its changing trend to obtain multiple pump efficiency sub-segments;

[0116] Calculate the trend parameters corresponding to each pump efficiency segment;

[0117] Based on multiple trend parameters, adjust the slope ratio and curve integral characteristics related to energy conversion efficiency in the dynamic response reference model to adjust the tolerance range.

[0118] In some embodiments, the efficiency curve of a water pump often exhibits complex trends, such as regions where efficiency increases, stabilizes, or decreases under different loads. If adjustments are made only based on the overall efficiency curve without fully considering these local trends, the dynamic response reference model may fail to accurately reflect the pump's true dynamic behavior under certain operating conditions, thus affecting the accuracy of dynamic behavior distortion analysis.

[0119] Therefore, the pump efficiency curve can be segmented to segment its changing trend, resulting in multiple pump efficiency sub-segments. The entire pump efficiency curve can be divided into several continuous sub-intervals based on its local variation characteristics. For example, different pump efficiency sub-segments can be identified based on changes in efficiency, the steepness of the slope, or the curvature. Each pump efficiency sub-segment represents the efficiency variation pattern of the pump within a specific operating range.

[0120] Then, the trend parameters corresponding to each pump efficiency segment are calculated. A quantitative analysis is performed on each identified pump efficiency segment to extract its core variation characteristics. These trend parameters may include, but are not limited to, the average slope, maximum slope, minimum slope, curvature, or the magnitude of efficiency change within the segment. The aim is to characterize the dynamic properties of the segment using a concise numerical form.

[0121] Then, based on multiple trend parameters, adjust the tolerance range of the slope ratio and curve integral feature related to energy conversion efficiency in the dynamic response reference model. This quantified trend information can be used to refine the parameters within the dynamic response reference model that are closely related to the pump's energy conversion efficiency. For example, when a pump efficiency segment shows a rapid increase in efficiency, the tolerance range of the slope ratio in the model can be appropriately tightened or adjusted to make it more sensitive to efficiency changes; conversely, when the efficiency is in a stable region, the tolerance range of the curve integral feature can be appropriately widened to accommodate smaller fluctuations.

[0122] To illustrate this technical solution more clearly, a specific example is used below. Suppose a water pump, during operation from low load to high load, exhibits an efficiency curve that first rises rapidly, reaches a peak, and then slowly declines. First, the efficiency curve is segmented to represent its changing trend. For example, three main efficiency segments can be identified: the first segment represents a rapid increase in efficiency, the second segment represents a stable peak efficiency region, and the third segment represents a slow decline in efficiency. Next, the trend parameter corresponding to each efficiency segment is calculated. For the first segment, the trend parameter might be a large positive slope value, indicating a rapid increase in efficiency; for the second segment, the trend parameter might be a slope value close to zero, indicating relatively stable efficiency; and for the third segment, the trend parameter might be a small negative slope value, indicating a slow decline in efficiency. Finally, based on these trend parameters, the tolerance range of the slope proportion and curve integral characteristics related to energy conversion efficiency in the dynamic response reference model is adjusted. Specifically, under the operating conditions corresponding to the first sub-segment, the tolerance range of the slope ratio in the model may be tightened to enhance the model's sensitivity to rapid changes in efficiency. Under the operating conditions corresponding to the second sub-segment, the tolerance range of the curve integral feature may be appropriately widened to allow the model to have a certain degree of robustness when efficiency is stable. Under the operating conditions corresponding to the third sub-segment, the model may be adjusted to better capture the dynamic characteristics when efficiency decreases slowly. In this way, the dynamic response reference model can adaptively adjust according to the actual operating characteristics of the pump in different efficiency regions, thereby enabling more accurate dynamic behavior distortion analysis.

[0123] Through the above technical solution, the dynamic response reference model in this embodiment can more accurately reflect the dynamic response characteristics of the water pump under different operating conditions. This refined adjustment based on the efficiency curve trend significantly improves the accuracy and reliability of dynamic behavior distortion analysis, avoiding misjudgments caused by the mismatch between the model and actual operating conditions. Therefore, it provides a more solid data foundation and decision-making basis for the adaptive scheduling of energy-saving water pump parameters, further enhancing the overall energy-saving effect and operational stability of the water pump system.

[0124] In some embodiments, step S204 involves subjecting the water pump to speed perturbation processing to obtain micro-perturbation response characteristics, which may include, but is not limited to, the following steps:

[0125] Initialize the speed disturbance parameters, which include the speed disturbance amplitude and the speed disturbance duration;

[0126] Calculate the pump efficiency and suction pressure based on the pump operating condition data and pump performance curve;

[0127] If the pump efficiency is less than the preset critical efficiency threshold, or the suction pressure is less than the preset cavitation risk pressure threshold, the speed disturbance parameters will be adjusted to reduce the speed disturbance amplitude and extend the speed disturbance duration.

[0128] Based on the adjusted speed disturbance parameters, the water pump is subjected to speed disturbance processing to obtain micro-disturbance response characteristics.

[0129] In some embodiments, if the pump's operating efficiency is too low or the suction pressure is close to the cavitation risk threshold, directly applying speed perturbation processing with preset amplitude and duration may lead to pump instability or even equipment damage risks such as cavitation, and may also fail to obtain accurate and effective micro-perturbation response characteristics. Therefore, speed perturbation parameters can be initialized first, including the speed perturbation amplitude and the speed perturbation duration. The speed perturbation amplitude refers to the range of pump speed variation from a reference speed, for example, it can be set to ±5% or ±10% of the reference speed. The speed perturbation duration refers to the length of time the pump speed remains in the perturbation state, for example, it can be set to several seconds to tens of seconds. These initial parameter settings aim to provide a benchmark for subsequent perturbation processing.

[0130] Then, based on the pump's operating data and performance curves, the pump efficiency and suction pressure are calculated. Pump performance curves describe the relationship between parameters such as head, flow rate, and efficiency at different pump speeds; these curves are typically provided by the pump manufacturer or obtained through on-site testing. Using this pump operating data and performance curves, the pump's current operating efficiency and suction pressure can be calculated in real time. Pump efficiency refers to the efficiency with which the pump converts input electrical energy into hydraulic energy, and suction pressure refers to the pressure at the pump inlet.

[0131] If the pump efficiency is less than the preset critical efficiency threshold, or the suction pressure is less than the preset cavitation risk pressure threshold, the speed disturbance parameters are adjusted to reduce the amplitude and duration of the speed disturbance. The preset critical efficiency threshold can be set based on the pump's design efficiency and actual operating experience; for example, it can be set to 70% or 80% of the pump's design efficiency. The preset cavitation risk pressure threshold refers to the lowest suction pressure that may cause cavitation in the pump. This threshold is usually related to the pump's net positive suction head (NPSH) characteristics and can be set based on those characteristics. When the calculated pump efficiency is less than the preset critical efficiency threshold, or the suction pressure is less than the preset cavitation risk pressure threshold, it indicates that the pump may be in an unfavorable operating state, requiring adjustment of the speed disturbance parameters. The adjustment strategy is to reduce the amplitude of the speed disturbance, for example, by halving it, and to extend the duration of the speed disturbance, for example, by doubling it. Reducing the disturbance amplitude aims to decrease the impact on pump operation and prevent further deterioration of the operating state or the initiation of cavitation under unfavorable conditions. Extending the duration of the perturbation helps to collect sufficiently stable perturbation response data even with smaller perturbation amplitudes.

[0132] Finally, based on the adjusted speed disturbance parameters, the pump speed is subjected to speed disturbance processing to obtain the micro-disturbance response characteristics. During the disturbance process, the pump speed will change according to the adjusted amplitude and duration, and the system will collect corresponding pressure response, power response, and flow response data to obtain the micro-disturbance response characteristics.

[0133] Through the above technical solution, this embodiment can significantly improve the safety, stability, and data acquisition effectiveness of the pump speed disturbance processing. By monitoring the pump efficiency and suction pressure in real time and adaptively adjusting the speed disturbance parameters accordingly, inappropriate disturbances can be effectively avoided when the pump is inefficient or at risk of cavitation, thereby protecting the pump equipment and extending its service life. Furthermore, while ensuring safety, extending the disturbance duration ensures that high-quality micro-disturbance response characteristics can be obtained even with small disturbance amplitudes, providing a more reliable data foundation for subsequent dynamic behavior distortion analysis, and thus improving the overall accuracy and robustness of the energy-saving pump parameter adaptive scheduling method.

[0134] In some embodiments, in step S205, the validity of the locally adjusted dynamic response reference model is verified based on the perturbation response characteristics to obtain the validity verification result, which may include, but is not limited to, the following steps:

[0135] Acquire sensor readings before and after the speed disturbance;

[0136] The sensor calibration drift is evaluated based on the sensor readings before and after the speed disturbance.

[0137] Monitor the proportional relationship between the rate of decrease in pipeline pressure and the rate of change in flow rate during speed disturbances;

[0138] By comparing the proportional relationship with historical normal leakage patterns, the pipeline leakage status can be identified;

[0139] Based on the pump efficiency curve, calculate the efficiency change before and after the speed disturbance;

[0140] The efficiency changes are compared with the dynamic response reference model after local adjustment to identify the degree of pump group efficiency decline;

[0141] Calculate the total deviation between the perturbation response characteristics and the dynamic response reference model after local adjustment;

[0142] Based on sensor calibration drift, pipeline leakage status, and pump efficiency degradation, the total deviation is decomposed to obtain multiple deviation components.

[0143] Based on multiple deviation components, it is determined whether the total deviation is caused by non-model failure factors, and the validity verification results are obtained.

[0144] In some embodiments, the dynamic behavior of a pump system may be affected by various non-model failure factors, such as sensor calibration drift, pipe leaks, or a decrease in the efficiency of the pump itself. If these factors are not fully considered, simply comparing the perturbation response characteristics with the model may lead to a misjudgment of the model's validity, thereby incorrectly triggering or canceling modifications to the pump parameters and affecting the accuracy and stability of energy-saving scheduling.

[0145] To this end, sensor readings before and after the speed disturbance can be obtained. Before and after the pump speed disturbance, key sensor data, such as the output values ​​of the pressure, flow, and power sensors, are recorded. These readings form the basis for subsequent system status evaluation. Sensor calibration drift is then assessed based on the sensor readings before and after the speed disturbance. This can be determined by comparing the differences in sensor readings under steady-state conditions before and after the disturbance, or by comparing them with known reference values. For example, if a significant and persistent deviation in sensor readings occurs under the same operating conditions (or through compensation calculations) before and after the disturbance, calibration drift can be considered to exist. The purpose is to eliminate the influence of sensor malfunctions or inaccuracies on the model validation results.

[0146] Then, the proportional relationship between the rate of pressure drop and the rate of flow change in the pipeline during the pump speed disturbance is monitored. During the pump speed disturbance, pipeline pressure and flow data are collected in real time, and their rates of change are calculated. By analyzing the dynamic proportional relationship between the rate of pressure drop and the rate of flow change, the hydraulic characteristics of the pipeline system can be reflected. The purpose is to provide crucial information for identifying pipeline leakage. The proportional relationship is then compared with historical normal leakage patterns to identify pipeline leakage. A historical database can be established, storing typical proportional relationships or ranges between the rate of pressure drop and the rate of flow change under normal, leak-free conditions. When the currently monitored proportional relationship deviates from this normal range, it may indicate a pipeline leak. The purpose is to distinguish between model failure and abnormal system behavior caused by pipeline leakage.

[0147] Next, based on the pump efficiency curve, calculate the efficiency change before and after the speed disturbance. Known or real-time calculated pump efficiency curves can be used, combined with operating parameters before and after the disturbance (such as flow rate, head, and power), to calculate the pump's operating efficiency change. The purpose is to quantify the pump set's own performance changes. The efficiency change is then compared with the dynamic response reference model after local adjustments to identify the degree of pump set efficiency decline. If the actual efficiency decline is significantly greater than the model's predicted value, it indicates that the pump set may have experienced efficiency degradation. The purpose is to assess the pump set's health status and eliminate interference from pump set performance degradation on model validation.

[0148] The total deviation between the perturbation response characteristics and the reference model of the dynamic response after local adjustments is calculated. This can be achieved by calculating statistical indicators such as mean square error and absolute error. The aim is to obtain a comprehensive index to measure the degree of fit between the model and the actual system behavior. The total deviation is decomposed into multiple deviation components based on sensor calibration drift, pipeline leakage status, and pump efficiency degradation. The total deviation can be attributed to different underlying causes. For example, if sensor drift exists, part of the total deviation can be attributed to sensor error; if pipeline leakage exists, another part of the deviation can be attributed to leakage. The purpose is to refine the analysis of deviation sources and avoid incorrectly attributing deviations caused by non-model factors to model failure.

[0149] Finally, based on multiple deviation components, it is determined whether the total deviation is caused by non-model failure factors, thus obtaining the validity verification result. After decomposing each deviation component, the contribution of these non-model factors (such as sensor drift, pipeline leakage, and pump efficiency reduction) to the total deviation is evaluated. If the main part of the total deviation can be explained by these non-model factors, the model itself is considered valid, but there are other problems with the system; conversely, if the contribution of non-model factors is small, but the total deviation is still large, it may indicate that the model does indeed have failures. The purpose is to obtain more accurate and reliable model validity verification conclusions.

[0150] Through the above technical solution, this embodiment can significantly improve the accuracy and robustness of dynamic response reference model validity verification. This embodiment can effectively distinguish between model failure and abnormal system behavior caused by non-model factors such as sensor malfunction, pipeline leakage, and pump performance degradation, avoiding unnecessary or erroneous adjustments to the model due to external interference. Therefore, it ensures the stability and reliability of the energy-saving pump parameter adaptive scheduling system under complex and variable operating conditions, preventing energy waste or system operation risks caused by misjudgment, thereby improving overall energy-saving effect and operating efficiency.

[0151] The beneficial effects of implementing the embodiments of the present invention include: First, the active power data of the frequency converter, pipeline pressure data, pipeline flow data, and motor speed data are collected and time-aligned. Then, response features are extracted from the time-aligned pump operating condition data to obtain dynamic response features. Then, dynamic behavior distortion analysis is performed based on the dynamic response reference model and dynamic response features to obtain dynamic behavior distortion analysis results. If the dynamic behavior distortion analysis results indicate the existence of behavior distortion, the modification of the pump parameters by the multiplicative correction coefficient is cancelled. This allows for the analysis of behavior distortion by combining pump operating condition data and dynamic response features, and the cancellation of pump parameter modification, thereby achieving adaptive scheduling of pump parameters, improving accuracy and stability, and reducing energy consumption per unit of water volume.

[0152] like Figure 2 As shown, this embodiment of the invention also provides an energy-saving water pump parameter adaptive scheduling system, including:

[0153] The data acquisition module 401 is used to acquire inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data;

[0154] The time alignment module 402 is used to perform time alignment processing on the inverter active power data, pipeline pressure data, pipeline flow data and motor speed data;

[0155] The response feature extraction module 403 is used to extract response features from the time-aligned pump operating condition data to obtain dynamic response features. The dynamic response features include response delay, rise slope, curve integral and response sequence. The pump operating condition data includes inverter active power data, pipeline pressure data, pipeline flow data or motor speed data.

[0156] The behavior distortion analysis module 404 is used to perform dynamic behavior distortion analysis based on the dynamic response reference model and dynamic response characteristics, and obtain the dynamic behavior distortion analysis results.

[0157] The parameter control module 405 is used to cancel the modification of the pump parameters by the multiplicative correction factor if the dynamic behavior distortion analysis result shows that behavior distortion exists.

[0158] The content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0159] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

Claims

1. An adaptive scheduling method for energy-saving water pump parameters, characterized in that, Includes the following steps: Collect inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data; Time alignment processing is performed on the active power data of the frequency converter, the pipeline pressure data, the pipeline flow data, and the motor speed data; Response features are extracted from the time-aligned water pump operating condition data to obtain dynamic response features. The dynamic response features include response delay, rise slope, curve integral, and response sequence. The water pump operating condition data includes the inverter active power data, the pipeline pressure data, the pipeline flow data, or the motor speed data. Based on the dynamic response reference model and the dynamic response characteristics, dynamic behavior distortion analysis is performed to obtain the dynamic behavior distortion analysis results. If the dynamic behavior distortion analysis result indicates the presence of behavior distortion, then the modification of the pump parameters by the multiplicative correction coefficient is cancelled; The process of extracting response features from the time-aligned pump operating data to obtain dynamic response features includes: Multi-level sliding median filtering is applied to the time-aligned pump operating data to remove instantaneous spikes and outliers. Adaptive threshold processing is applied to the pump operation data after multi-layer sliding median filtering to identify normal and abnormal data segments. Calculate the response delay, the rise slope, and the curve integral based on the normal data segment; Identify adjacent data segments that are adjacent to the abnormal data segment; Based on the adjacent data segments, the abnormal data segments are subjected to smooth interpolation to obtain a smooth response curve; Extract the response order from the smooth response curve.

2. The method of claim 1, wherein, The step of performing dynamic behavior distortion analysis based on the dynamic response reference model and the dynamic response characteristics to obtain dynamic behavior distortion analysis results includes: Based on the pump operating condition data, identify the pump operating status; If the water pump is in a stable operating state, then the water pump efficiency curve is calculated based on the water pump operating condition data. Based on the pump efficiency curve, the dynamic response reference model is locally adjusted. During low-load periods, the pump speed is perturbed to obtain micro-perturbation response characteristics, which include pressure response, power response and flow response. Based on the perturbation response characteristics, the effectiveness of the dynamic response reference model after local adjustment is verified, and the effectiveness verification results are obtained. Based on the validity verification results, the dynamic response characteristics, and the dynamic response reference model after local adjustment, dynamic behavior distortion analysis is performed to obtain the dynamic behavior distortion analysis results.

3. The method according to claim 2, characterized in that, The step of identifying the water pump operating status based on the water pump operating condition data includes: Based on the pump operating condition data, calculate the average data fluctuation amplitude, average value, and standard deviation within the sliding time window; If the current data fluctuation amplitude is greater than the average data fluctuation amplitude, then the water pump operating state is determined to be unstable; If the current data fluctuation range is less than or equal to the average data fluctuation range, then the average value and the standard deviation are judged. If the average value is less than a preset average threshold and the standard deviation is less than a preset deviation threshold, then monitor the grid voltage data and the upstream main pipe pressure data. The operating status of the water pump is identified based on the grid voltage data and the upstream main pipe pressure data.

4. The method of claim 3, wherein, The step of identifying the operating status of the water pump based on the grid voltage data and the upstream main pipe pressure data includes: Harmonic analysis is performed on the power grid voltage data to extract the harmonic distortion rate and instantaneous voltage drop amplitude; Frequency analysis was performed on the upstream main pipe pressure data to identify periodic pulsations and sudden changes; Based on the pump operating data, calculate the instantaneous rate of change and the trend of change; Based on the harmonic distortion rate, the instantaneous drop amplitude, and the instantaneous change rate, anomaly analysis is performed on the power grid voltage data to obtain the first anomaly analysis result; Based on the periodic pulsations, sudden changes, and trends, anomaly analysis is performed on the upstream main pipe pressure data to obtain a second anomaly analysis result. If both the first anomaly analysis result and the second anomaly analysis result are normal, then the pump's operating state is determined to be stable; otherwise, the pump's operating state is determined to be pseudo-stable.

5. The method of claim 2, wherein, The step of calculating the pump efficiency curve based on the pump operating condition data includes: The pump operating data is processed by time series segmentation to obtain complete data segments, insufficient data segments, and unevenly distributed data segments. The first efficiency curve is obtained by fitting the complete data segment. Based on historical operating data, data supplementation is performed on the data segments with insufficient data volume; The second efficiency curve is obtained by fitting the insufficient data segment after data supplementation. The unevenly distributed data segment was fitted using a locally weighted linear regression method to obtain a third efficiency curve; The first efficiency curve, the second efficiency curve, and the third efficiency curve are smoothed by interpolation to generate the pump efficiency curve.

6. The method of claim 2, wherein, The step of performing local adjustments to the dynamic response reference model based on the pump efficiency curve includes: The pump efficiency curve is segmented according to its changing trend to obtain multiple pump efficiency sub-segments; Calculate the trend parameters corresponding to each pump efficiency segment; Based on multiple trend parameters, the tolerance range of the slope ratio and curve integral characteristics related to energy conversion efficiency in the dynamic response reference model is adjusted.

7. The method of claim 2, wherein, The process of perturbing the pump's rotational speed to obtain micro-perturbation response characteristics includes: Initialize the speed disturbance parameters, which include the speed disturbance amplitude and the speed disturbance duration; Based on the pump operating condition data and pump performance curve, calculate the pump efficiency and suction pressure; If the pump efficiency is less than a preset critical efficiency threshold, or the suction pressure is less than a preset cavitation risk pressure threshold, the speed disturbance parameter is adjusted to reduce the speed disturbance amplitude and prolong the speed disturbance duration. Based on the adjusted speed disturbance parameters, the water pump is subjected to speed disturbance processing to obtain the micro-disturbance response characteristics.

8. The method of claim 2, wherein, The step of validating the dynamic response reference model after local adjustment based on the perturbation response characteristics, and obtaining the validity verification result, includes: Acquire sensor readings before and after the speed disturbance; The sensor calibration drift is evaluated based on the sensor readings before and after the speed disturbance. Monitor the proportional relationship between the rate of decrease in pipeline pressure and the rate of change in flow rate during speed disturbances; The aforementioned ratio is compared with historical normal leakage patterns to identify pipeline leakage status; Based on the pump efficiency curve, calculate the efficiency change before and after the speed disturbance; The efficiency change is compared with the dynamic response reference model after local adjustment to identify the degree of pump group efficiency decline. Calculate the total deviation between the perturbation response characteristics and the dynamic response reference model after local adjustment; Based on the sensor calibration drift, the pipeline leakage status, and the degree of efficiency reduction of the pump set, the total deviation is decomposed to obtain multiple deviation components; Based on the multiple deviation components, it is determined whether the total deviation is caused by non-model failure factors, and the validity verification result is obtained.

9. An energy-saving water pump parameter adaptive scheduling system, used to execute the method as described in any one of claims 1-8, characterized in that, include: The data acquisition module is used to collect inverter active power data, pipeline pressure data, pipeline flow data, and motor speed data; The time alignment module is used to perform time alignment processing on the inverter active power data, the pipeline pressure data, the pipeline flow data, and the motor speed data; The response feature extraction module is used to extract response features from the time-aligned pump operating condition data to obtain dynamic response features. The dynamic response features include response delay, rise slope, curve integral, and response sequence. The pump operating condition data includes the inverter active power data, the pipeline pressure data, the pipeline flow data, or the motor speed data. The behavior distortion analysis module is used to perform dynamic behavior distortion analysis based on the dynamic response reference model and the dynamic response characteristics, and obtain the dynamic behavior distortion analysis results. The parameter control module is used to cancel the modification of the pump parameters by the multiplicative correction coefficient if the dynamic behavior distortion analysis result indicates the existence of behavior distortion.

Images