A water pump operation efficiency monitoring and abnormal early warning threshold self-adaptive setting method

Abstract

This invention relates to the field of water pump operation monitoring technology, and discloses a method for adaptively setting water pump operating efficiency monitoring and anomaly early warning thresholds. The method first collects operating parameters of the target water pump within a preset period to determine its real-time operating status; then, based on this, it divides the data into efficiency-dominant and secondary collection points, and combines these with environmental interference monitoring points to form a set of monitoring points to be analyzed; subsequently, it calculates the parameter influence coefficients of related monitoring points in the set, constructing a parameter association network with monitoring points as nodes and influence coefficients as weights; next, it determines the core path of the network and prioritizes the monitoring points; finally, it processes the data according to priority to achieve adaptive setting of the anomaly early warning threshold. This method, through dynamic division of monitoring points, analysis of parameter associations, and sorting processing, makes the threshold setting more closely match the actual operating status of the water pump, improving the targeting and adaptability of monitoring.

Description

Technical Field

[0001] This invention relates to the field of water pump operation monitoring technology, specifically a method for adaptively setting water pump operation efficiency monitoring and abnormal early warning thresholds. Background Technology

[0002] In many fields such as industrial production, municipal water supply, and agricultural irrigation, water pumps are core equipment for fluid transportation, and their operating efficiency directly affects energy consumption, system stability, and operating costs. With the development of intelligent technology, real-time monitoring and early warning of anomalies in water pump operation have become an important part of equipment management.

[0003] Currently, pump operating efficiency monitoring largely relies on fixed parameter collection points. The collected parameters typically include flow rate, head, power, speed, and temperature. These parameters are used to evaluate the real-time efficiency of the pump. However, in practical applications, due to the complex and variable operating environment of the pump, such as changes in water quality, fluctuations in pipeline pressure, and differences in ambient temperature, the monitoring data from fixed collection points often fails to fully reflect the pump's true operating status.

[0004] Existing abnormal warning threshold settings mostly rely on empirical values ​​or fixed thresholds, lacking adaptability to dynamic changes in pump operating conditions. When a pump is under different operating conditions (such as startup, shutdown, and load changes), the range of its normal operating parameters will change significantly. Fixed thresholds can easily lead to inaccurate warnings, resulting in false alarms or missed alarms. For example, during the pump startup phase, power and speed will rise sharply in a short period of time. If the thresholds used during normal operation are applied, false alarms are very likely to be triggered. Conversely, during long-term low-load operation, the slow deviation of some parameters may be ignored by fixed thresholds, resulting in potential faults not being detected in time.

[0005] Existing monitoring methods do not adequately consider the correlation between parameters, often analyzing each parameter as an independent entity and ignoring their mutual influence. For example, changes in flow rate may cause corresponding changes in head and power. If a threshold is set for a single parameter, it is difficult to accurately determine the root cause of the anomaly, increasing the difficulty of troubleshooting. These problems reduce the accuracy and reliability of pump operating efficiency monitoring, affecting the timeliness and effectiveness of equipment maintenance. Summary of the Invention

[0006] The purpose of this invention is to provide a method for adaptively setting the threshold for monitoring the operating efficiency of water pumps and for issuing early warnings of abnormalities, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides a method for adaptively setting a threshold for monitoring and issuing early warning of abnormalities in water pump operation, the method comprising:

[0008] Collect various operating parameters of the currently running target water pump within a preset period, and determine the real-time operating status of the target water pump based on the operating parameters;

[0009] Based on the real-time operating status, the operating parameter collection points are divided into efficiency-dominant collection points and efficiency-secondary collection points. Environmental interference monitoring points are determined from the auxiliary monitoring points other than the operating parameter collection points. The efficiency-dominant collection points, efficiency-secondary collection points, and environmental interference monitoring points constitute the set of monitoring points to be analyzed.

[0010] Determine the parameter influence coefficient between any two related monitoring points in the set of monitoring points to be analyzed, and construct the parameter association network of the set of monitoring points to be analyzed, using the monitoring points in the set of monitoring points to be analyzed as nodes and the parameter influence coefficient as the weight between the connecting nodes.

[0011] Determine the core path of the parameter association network, and based on the core path, determine the priority order of each monitoring point in the set of monitoring points to be analyzed;

[0012] The data of each monitoring point in the set of monitoring points to be analyzed are processed sequentially according to the priority order, so as to achieve adaptive setting of the abnormal early warning threshold.

[0013] Preferably, determining the real-time operating status of the target water pump based on the operating parameters includes:

[0014] Based on the operating parameters, calculate the efficiency benchmark values ​​of the target water pump over multiple historical periods;

[0015] Based on the fluctuation range of the operating parameters corresponding to each efficiency benchmark value, the calculated efficiency benchmark values ​​are grouped to form one or more efficiency state groups.

[0016] A target efficiency state group that matches the current operating parameters of the target water pump is determined from the efficiency state group, and the state identifier corresponding to the target efficiency state group is used as the real-time operating state of the target water pump.

[0017] Preferably, the calculated efficiency benchmark values ​​are grouped according to the fluctuation range of the operating parameters corresponding to each efficiency benchmark value, including:

[0018] Obtain the parameter fluctuation range preset in the water pump monitoring system;

[0019] For any calculated efficiency benchmark value, identify the parameter fluctuation range to which the fluctuation range of the operating parameters corresponding to the efficiency benchmark value belongs, and classify the efficiency benchmark value into the efficiency state group corresponding to the identified parameter fluctuation range.

[0020] Preferably, based on the real-time operating status, the operating parameter collection points are divided into efficiency-dominant collection points and efficiency-secondary collection points, including:

[0021] Determine the key and non-key impact parameters corresponding to the real-time operating status;

[0022] A first monitoring range for the critical impact parameter is determined at the operating parameter acquisition points, and a second monitoring range for the non-critical impact parameter is determined at the operating parameter acquisition points.

[0023] The operating parameter collection points covered by the first monitoring range are determined as efficiency-dominant collection points, and the operating parameter collection points covered by the second monitoring range are determined as efficiency-secondary collection points.

[0024] Preferably, the environmental interference monitoring points identified among the auxiliary monitoring points other than the operating parameter acquisition points include:

[0025] A first influence radius is set for the efficiency-dominant collection point among the operating parameter collection points, and a second influence radius is set for the efficiency-secondary collection point among the operating parameter collection points;

[0026] For auxiliary monitoring points other than the operating parameter collection points, if the distance to the efficiency-dominant collection point is within the first influence radius, then the auxiliary monitoring point is selected as the environmental interference monitoring point; if the distance to the efficiency-secondary collection point is within the second influence radius, then the auxiliary monitoring point is selected as the environmental interference monitoring point.

[0027] Preferably, determining the parameter influence coefficient between any two related monitoring points in the set of monitoring points to be analyzed includes:

[0028] For the first and second monitoring points associated in the set of monitoring points to be analyzed, the parameter types of the first and second monitoring points are identified respectively. The parameter types include one of the following: efficiency dominant parameter, efficiency secondary parameter, and environmental disturbance parameter.

[0029] Based on the identified parameter types, determine the first influence weight of the first monitoring point and the second influence weight of the second monitoring point;

[0030] Calculate the correlation between the first influence weight and the historical data of the second monitoring point, and determine the parameter influence coefficient between the first monitoring point and the second monitoring point based on the correlation.

[0031] Preferably, determining the parameter influence coefficient between the first monitoring point and the second monitoring point based on the correlation includes:

[0032] Collect historical operating data of each monitoring point in the set of monitoring points to be analyzed, and determine the variance of the collected historical operating data;

[0033] Based on the volatility variance, multiple correlation levels are defined, and the target correlation level of the calculated first influence weight and the correlation of historical data is determined.

[0034] The preset coefficient value corresponding to the target correlation level is determined as the parameter influence coefficient between the first monitoring point and the second monitoring point.

[0035] Preferably, processing the data of each monitoring point in the set of monitoring points to be analyzed sequentially according to the priority order to achieve adaptive setting of the anomaly warning threshold includes:

[0036] Real-time parameter data of each monitoring point in the priority ranking is extracted, and a time-series feature sequence is constructed based on the extracted real-time parameter data. The arrangement order of each feature in the time-series feature sequence is consistent with the priority ranking.

[0037] The time-series feature sequence is input into an adaptive threshold model to generate an anomaly warning threshold for each monitoring point.

[0038] Monitoring is performed on the corresponding monitoring points according to the generated abnormal warning thresholds in order to achieve abnormal warning.

[0039] Preferably, the method further includes:

[0040] After the abnormal warning threshold is set, the parameter change rate of each monitoring point in the set of monitoring points to be analyzed is continuously monitored.

[0041] When the rate of change of the parameter exceeds the preset fluctuation threshold for multiple consecutive periods, the step of determining the core path of the parameter association network is re-executed, and the priority sorting is adjusted according to the new core path.

[0042] Preferably, determining the key and non-key impact parameters corresponding to the real-time operating status includes:

[0043] Obtain the design parameter table of the target water pump, which includes the rated flow rate, rated head, and rated power.

[0044] The actual operating parameters in the real-time operating status are compared with the rated parameters in the design parameter table, and the percentage of parameter deviation is calculated.

[0045] Operating parameters whose parameter deviation percentage is greater than a preset deviation threshold are identified as critical influencing parameters, while operating parameters whose parameter deviation percentage is less than or equal to the preset deviation threshold are identified as non-critical influencing parameters.

[0046] Compared with the prior art, the beneficial effects of the present invention are:

[0047] By dynamically dividing the operational parameter collection points into efficiency-dominant and efficiency-secondary collection points, and combining them with environmental interference monitoring points, a set of monitoring points to be analyzed is formed, making the monitoring scope more closely reflect the actual operating state of the water pump. This division method is not fixed but is adjusted based on the real-time operating status. It can focus on parameters that have a more significant impact on efficiency according to the characteristics of the water pump under different operating conditions, while also taking into account environmental factors that may cause interference, making the monitoring more targeted.

[0048] In constructing a parameter correlation network, the intrinsic relationships between parameters can be clearly presented by calculating the parameter influence coefficient between any two related monitoring points and using this coefficient as the weight to connect nodes. This network structure no longer treats parameters in isolation but analyzes their interactions from a holistic perspective. For example, changes in ambient temperature may affect the power parameters of a water pump, and changes in power may be related to fluctuations in flow rate and head. The network can intuitively reflect these chain reactions, providing a more comprehensive perspective for subsequent analysis.

[0049] Determining the core path of the parameter association network and prioritizing monitoring points makes data processing more logical and hierarchical. Processing data sequentially according to priority allows for focusing on parameters more critical to efficiency and anomaly warning, reducing interference from irrelevant data. This orderly processing approach avoids analytical chaos caused by information overload and makes the setting of anomaly warning thresholds more systematic.

[0050] The adaptive setting of the abnormal warning threshold can dynamically adjust according to changes in the water pump's operating status. When the water pump is in different operating conditions such as startup, high load, and low load, the threshold will change accordingly, thus better conforming to the normal parameter range under each operating condition. This adaptability enables the warning to more accurately identify true abnormalities, reduce false judgments caused by fixed thresholds, and also promptly capture potential problems under different operating conditions, improving the ability to perceive the water pump's operating status. Attached Figure Description

[0051] Figure 1 This is a schematic diagram illustrating the working principle of the water pump operating efficiency monitoring and abnormal early warning threshold adaptive setting method described in this invention.

[0052] Figure 2 A flowchart for determining the real-time operating status;

[0053] Figure 3 Flowchart for determining the influence coefficients of parameters;

[0054] Figure 4 A detailed flowchart for calculating the parameter influence coefficients;

[0055] Figure 5 The flowchart for determining critical and non-critical impact parameters. Detailed Implementation

[0056] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] Please see Figure 1 This invention provides a method for adaptively setting a threshold for monitoring and warning of abnormalities in water pump operation efficiency, the method comprising:

[0058] Collect various operating parameters of the currently running target water pump within a preset period, and determine the real-time operating status of the target water pump based on the operating parameters.

[0059] The preset cycle can be set according to the operating characteristics of the water pump, for example, 30 minutes per cycle. Operating parameters include, but are not limited to, flow rate, head, power, speed, inlet and outlet pressure, and medium temperature. These parameters are collected by sensors deployed at different locations on the water pump. After collection, the parameters are preprocessed to remove outliers and noise data, and then the real-time operating status of the water pump is determined based on the processed parameters.

[0060] Based on the real-time operating status, the operating parameter collection points are divided into efficiency-dominant collection points and efficiency-secondary collection points. Environmental interference monitoring points are determined from the auxiliary monitoring points other than the operating parameter collection points. The efficiency-dominant collection points, efficiency-secondary collection points, and environmental interference monitoring points constitute the set of monitoring points to be analyzed.

[0061] Real-time operating status reflects the current efficiency level and operating mode of the water pump. Based on this, efficiency-dominant data collection points are those that collect parameters that have a significant impact on pump efficiency, such as the installation points of power sensors and flow sensors. Efficiency-secondary data collection points are those that collect parameters that have a smaller impact on efficiency, such as the installation point of the medium temperature sensor. Auxiliary monitoring points include monitoring points for the temperature, humidity, and vibration of the water pump's surrounding environment. From these, environmental interference monitoring points that may interfere with the pump's operating parameters are selected, such as ambient temperature monitoring points located close to the pump. These three types of monitoring points together form the set of monitoring points to be analyzed.

[0062] Determine the parameter influence coefficient between any two related monitoring points in the set of monitoring points to be analyzed, and construct a parameter association network for the set of monitoring points to be analyzed, using the monitoring points in the set of monitoring points to be analyzed as nodes and the parameter influence coefficient as the weight between the connecting nodes.

[0063] Related monitoring points refer to monitoring points where parameters influence each other. For example, power acquisition points and flow acquisition points are related, as are ambient temperature acquisition points and medium temperature acquisition points. By calculating the parameter influence coefficients between these related monitoring points, the degree of influence between parameters is quantified. Then, using monitoring points as nodes and influence coefficients as weights, a parameter association network is constructed using graph theory. The weight value of each edge in the network is the corresponding parameter influence coefficient.

[0064] The core path of the parameter association network is determined, and the priority order of each monitoring point in the set of monitoring points to be analyzed is determined according to the core path.

[0065] The core path refers to the path in the parameter association network that has the greatest impact on pump efficiency. It can be determined by analyzing the connection strength and weight values ​​of the nodes in the network, for example, by using the maximum weight path algorithm to select the core path. Based on the importance of the monitoring points on the core path, combined with the correlation between other monitoring points and the core path, all monitoring points to be analyzed are prioritized. Key nodes on the core path have higher priority, while nodes with weaker correlation to the core path have lower priority.

[0066] The data of each monitoring point in the set of monitoring points to be analyzed are processed sequentially according to the priority order, so as to achieve adaptive setting of the abnormal early warning threshold.

[0067] Data from each monitoring point is analyzed and processed sequentially in descending order of priority, with priority given to data from high-priority monitoring points. Then, based on the processing results and the correlation between parameters, the abnormal warning thresholds for each monitoring point are dynamically adjusted to adapt to changes in the pump's operating status.

[0068] Example 1: See Figure 2When determining the real-time operating status of the target water pump based on operating parameters, the efficiency benchmark value of the target water pump is calculated over multiple historical periods based on these parameters. The efficiency benchmark value is derived from the ratio of the effective power generated during pump operation to the input power. The effective power needs to be calculated comprehensively by combining real-time collected flow rate, head, and density of the transported medium. The input power can be directly obtained from the power sensor installed on the pump motor, or derived from collected current and voltage data. The selection of historical periods needs to cover different operating conditions of the water pump; the number can be set according to actual monitoring needs, for example, selecting the most recent 60 periods, each corresponding to one hour of continuous operating data, to ensure that the calculated efficiency benchmark value can reflect the pump's operating efficiency characteristics over different time periods. During the calculation process, the collected operating parameters need to be preprocessed to remove abnormal data caused by sensor failure or transient interference. Missing data is supplemented using interpolation to ensure data integrity and continuity, thereby improving the accuracy of the efficiency benchmark value calculation.

[0069] Based on the fluctuation range of operating parameters corresponding to each efficiency benchmark value, the calculated efficiency benchmark values ​​are grouped to form one or more efficiency status groups. In practice, the parameter fluctuation ranges preset in the pump monitoring system are first obtained. These ranges are defined based on the pump's design parameters, industry standards, and long-term operational experience data. Different operating parameters correspond to different fluctuation ranges. For example, the fluctuation range of flow parameters can be divided into multiple levels, and parameters such as head, power, and speed are also set with corresponding fluctuation ranges according to their impact on pump efficiency. The division of these ranges must consider the natural fluctuation range of parameters during normal pump operation, while also taking into account the reasonable variation range of parameters under different operating conditions, to ensure that the grouping results accurately reflect the pump's efficiency status.

[0070] For any calculated efficiency benchmark value, identify the parameter fluctuation range to which the corresponding operating parameter fluctuation range belongs, and assign the efficiency benchmark value to the efficiency state group corresponding to the identified parameter fluctuation range. Each efficiency benchmark value corresponds to a set of operating parameter fluctuation data, and it is necessary to analyze which preset fluctuation range these parameters belong to one by one. For example, if the flow fluctuation range corresponding to a certain efficiency benchmark value is ±4%, and the preset flow fluctuation range in the system is ±5% or less, then the flow fluctuation range belongs to this range; similarly, if the head fluctuation range corresponding to the efficiency benchmark value is ±7%, and the head fluctuation range is ±10% or less, then the head fluctuation range belongs to this range. If the combination of the ±5% flow fluctuation range and the ±10% head fluctuation range in the system corresponds to an efficiency state group, then the efficiency benchmark value will be assigned to that group. In the case of multiple operating parameters, it is necessary to comprehensively consider the fluctuation range assignment of all parameters to determine the efficiency state group to which the efficiency benchmark value belongs, ensuring that each efficiency benchmark value can be accurately assigned to the corresponding group.

[0071] Within the efficiency status groups, a target efficiency status group matching the current operating parameters of the target pump is identified, and the status identifier corresponding to the target efficiency status group is used as the real-time operating status of the target pump. The acquisition and processing of real-time operating parameters are consistent with historical parameters, requiring preprocessing to remove interference. The fluctuation range of the current operating parameters is compared with the parameter fluctuation ranges corresponding to each efficiency status group. When the fluctuation range of the current parameters completely matches all parameter fluctuation ranges corresponding to a certain efficiency status group, that group is determined as the target efficiency status group. The status identifier is a characteristic description of the efficiency status group, such as "high-efficiency stable operation," "medium-efficiency fluctuating operation," and "low-efficiency operation." These identifiers intuitively reflect the current operating status of the pump, providing a basis for subsequent parameter acquisition point division and abnormal warning threshold setting. During the matching process, if the fluctuation range of the current operating parameters simultaneously meets the range requirements of multiple efficiency status groups, it is necessary to further combine the parameter change trend and historical matching records to select the efficiency status group that best matches the current operating conditions as the target group, ensuring the accuracy of the real-time operating status determination.

[0072] Example 2: See Figure 5 Based on real-time operating status, when dividing the operating parameter collection points into efficiency-dominant and efficiency-secondary collection points, the first step is to determine the key and non-key influencing parameters corresponding to the real-time operating status. The design parameter table of the target water pump is obtained. This table, provided by the water pump manufacturer, contains various rated parameters of the water pump under standard operating conditions. In addition to rated flow rate, rated head, and rated power, it may also include rated speed, rated inlet and outlet pressure, etc. These parameters together constitute the benchmark reference for the normal operation of the water pump.

[0073] The actual operating parameters in real-time are compared with the rated parameters in the design parameter table to analyze the differences. The actual operating parameters are collected in real-time by sensors deployed at key parts of the pump, covering data such as flow rate, head, and power corresponding to the rated parameters. During the comparison process, it is necessary to ensure that the acquisition conditions for the actual operating parameters and rated parameters are consistent. For example, environmental conditions such as medium temperature and density must conform to the standard conditions marked in the design parameter table. If discrepancies exist, the actual operating parameters must be corrected to ensure the reliability of the comparison results.

[0074] After comparing the differences between parameters, operating parameters with a percentage deviation greater than a preset deviation threshold are identified as critical influencing parameters, while those with a percentage deviation less than or equal to the preset deviation threshold are identified as non-critical influencing parameters. The preset deviation threshold is determined based on the type of water pump, application scenario, and operating requirements. Different parameters may have different deviation thresholds. For example, for water pumps used in precision water supply systems, the deviation threshold for flow rate parameters may be set relatively low, while for water pumps used for general irrigation, the deviation threshold for power parameters may be set relatively high.

[0075] After determining the critical and non-critical influencing parameters, a first monitoring range for the critical influencing parameters and a second monitoring range for the non-critical influencing parameters are determined at the operational parameter acquisition points. The first monitoring range refers to the installation locations and coverage areas of all sensors capable of acquiring critical influencing parameters. These locations are typically distributed in critical areas that significantly impact pump efficiency. For example, a sensor acquiring flow parameters might be installed on the pump's outlet pipe, and a sensor acquiring power parameters might be integrated into the motor's control circuit. These sensor locations collectively constitute the first monitoring range. The second monitoring range consists of the locations and coverage areas of sensors acquiring non-critical influencing parameters. For example, a sensor monitoring medium temperature might be installed on the pump's inlet pipe, and a sensor monitoring ambient humidity might be installed inside the pump room. These locations form the second monitoring range.

[0076] The operating parameter collection points covered by the first monitoring range are designated as efficiency-dominant collection points, and those covered by the second monitoring range are designated as efficiency-secondary collection points. Data collected from efficiency-dominant collection points directly reflects the core influencing factors of pump efficiency, and its data quality and real-time performance are crucial for subsequent efficiency monitoring and anomaly early warning. During data processing, data from these collection points will be prioritized for processing and analysis to ensure timely detection of anomalies that may affect pump efficiency. Data collected from efficiency-secondary collection points has a relatively smaller impact on pump efficiency, and its data processing priority is lower than that of efficiency-dominant collection points. However, it is not negligible; this data provides supplementary information for a comprehensive assessment of the pump's operating status, helping to more accurately determine the overall pump's operating condition. This classification method enables the pump monitoring system to process various monitoring data more effectively, improving system operating efficiency and monitoring accuracy.

[0077] Example 3: See Figure 3 When determining environmental interference monitoring points among auxiliary monitoring points outside of the operating parameter acquisition points, a first influence radius is first set for the efficiency-dominant acquisition points among the operating parameter acquisition points, and a second influence radius is set for the efficiency-secondary acquisition points among the operating parameter acquisition points. The values ​​of the first and second influence radii need to be determined in conjunction with the physical characteristics of the parameters and the actual monitoring scenario. For example, for acquisition points collecting efficiency-dominant parameters such as power and flow rate, which are more sensitive to environmental vibration, the first influence radius can be set appropriately larger to cover areas that may cause significant vibration interference; while for acquisition points collecting efficiency-secondary parameters such as medium temperature, which are less affected by the environment, the second influence radius can be set smaller. The setting of these radii needs to be based on on-site investigation of the pump's operating environment, taking into account factors such as the distribution of surrounding equipment and wall obstructions, to ensure that environmental factors that may interfere with the corresponding parameters can be accurately captured.

[0078] For auxiliary monitoring points other than the operational parameter acquisition points, if the distance to the efficiency-dominant acquisition point is within the first radius of influence, the auxiliary monitoring point is selected as an environmental interference monitoring point; if the distance to the efficiency-secondary acquisition point is within the second radius of influence, the auxiliary monitoring point is selected as an environmental interference monitoring point. Auxiliary monitoring points are diverse, including ambient temperature and humidity sensors installed in the pump room, vibration sensors installed near the equipment, and noise sensors installed on pipe supports. During the selection process, the straight-line distance between each auxiliary monitoring point and each efficiency-dominant and efficiency-secondary acquisition point must be measured individually. The distance can be calculated using spatial distance calculation methods based on the location coordinates recorded during installation. When the distance between an auxiliary monitoring point and any efficiency-dominant acquisition point is less than or equal to the first radius of influence, the auxiliary monitoring point is included in the environmental interference monitoring point; similarly, if the distance to any efficiency-secondary acquisition point is less than or equal to the second radius of influence, it will also be identified as an environmental interference monitoring point. This method ensures that all environmental factors that may interfere with operational parameters are included in the monitoring scope.

[0079] When determining the parameter influence coefficient between any two related monitoring points in the set of monitoring points to be analyzed, for the first and second related monitoring points in the set of monitoring points to be analyzed, the parameter types of the first and second monitoring points are identified respectively. Parameter types include one of the following: efficiency-dominant parameters, efficiency-secondary parameters, and environmental interference parameters. The identification of parameter types is based on the attributes of the monitoring points and the nature of the collected parameters. For example, the flow parameter collected by the flow sensor installed on the water pump outlet pipe is an efficiency-dominant parameter; the medium temperature parameter collected by the temperature sensor installed on the inlet pipe is an efficiency-secondary parameter; and the parameter collected by the ambient temperature sensor installed in the machine room is an environmental interference parameter. During the identification process, it is necessary to refer to the preset parameter type classification table to ensure that the parameter type of each monitoring point is accurately classified, providing a basis for the subsequent determination of influence weights.

[0080] Based on the identified parameter types, the first influence weight for the first monitoring point and the second influence weight for the second monitoring point are determined. The numerical values ​​for the first and second influence weights are set based on the physical impact mechanism of the monitoring point parameter types on pump operating efficiency and industry-recognized parameter importance grading rules. The specific relationships are as follows:

[0081] If the monitoring point parameter type is an efficiency-dominant parameter (such as power, flow rate, head), its corresponding influence weight (including the first or second influence weight) should be set to the highest level. This type of parameter directly determines the ratio of the pump's effective power to its input power and is a core indicator reflecting the pump's efficiency. For example, the power parameter is directly related to the conversion efficiency of motor energy consumption and output power, and the flow rate parameter directly affects the fluid transport efficiency. Therefore, it needs to be given the highest weight to reflect its dominant role in efficiency.

[0082] If the monitoring point parameter type is a secondary efficiency parameter (such as medium temperature, inlet and outlet pressure), its corresponding influence weight is set to medium level. This type of parameter affects pump efficiency indirectly. For example, changes in medium temperature will cause changes in fluid density, which in turn affects the effective power calculation result. However, compared with parameters such as power and flow rate, its influence on efficiency is weaker, so its weight level is lower than that of the primary efficiency parameter.

[0083] If the monitoring point parameter type is an environmental interference parameter (such as ambient temperature, machine room humidity, and surrounding vibration), its corresponding influence weight is set to the lowest level. This type of parameter indirectly affects the pump's operating status through external environmental factors. For example, excessively high ambient temperature may cause a decrease in the motor's heat dissipation efficiency, indirectly affecting power consumption. However, it needs to be transmitted through multiple links to affect efficiency, resulting in a long influence path and a weak degree of influence. Therefore, its weight level is the lowest.

[0084] Weighting rules: The influence weights for parameters of the same type remain consistent. The weights for different parameter types satisfy the relationship: "Weight of primary efficiency parameter > Weight of secondary efficiency parameter > Weight of environmental interference parameter." The specific values ​​for these weights must be determined based on the target pump's equipment type (e.g., centrifugal pump, axial flow pump), application scenario (e.g., industrial water supply, agricultural irrigation), and design standards, using the following methods:

[0085] Refer to the annotations on the importance of parameters in the pump design manual. For example, if the design manual for a certain type of centrifugal pump clearly states that "flow rate and head are the core parameters affecting efficiency, while medium temperature is a secondary parameter affecting efficiency", then match the corresponding weight level according to the annotation.

[0086] Based on the requirements for parameter monitoring priority in industry technical specifications, the weight level is derived by reverse calculation according to the monitoring priority, with higher weight corresponding to high-frequency monitoring parameters;

[0087] Based on historical operating data of similar water pumps, if the historical data shows that the correlation between changes in a certain type of parameter (such as power) and changes in efficiency is the most frequent and the largest, then the rationality of the corresponding weight level of that type of parameter is confirmed.

[0088] The magnitude of the influence weight reflects the degree of influence of the parameter type on the pump efficiency. The determination of these weights should refer to the pump design manual, industry standards, and the correlation between the parameters and efficiency changes in historical operating data. They should be set after comprehensive analysis, and the influence weight of parameters of the same type should be kept consistent to ensure the uniformity of the analysis.

[0089] The correlation between the historical data of the first and second monitoring points is calculated to determine the parameter influence coefficient between them. The historical data correlation is calculated based on parameter data from the two monitoring points over multiple historical periods, using statistical analysis to determine the degree of correlation in their changing trends. During the calculation, the historical data must first be standardized to eliminate the influence of different parameter units and magnitudes, making the correlation analysis more comparable.

[0090] The parameter influence coefficient can be determined using the following formula:

[0091]

[0092] in, This represents the parameter influence coefficient between the first and second monitoring points. This indicates the first influence weight of the first monitoring point. This represents the correlation value between the first influence weight and the historical data of the second monitoring point.

[0093] The correlation value ranges from 0 to 1, with a higher value indicating a stronger correlation. In actual calculations, a sufficient amount of historical data needs to be collected, typically the data from the most recent 100 periods, to ensure the reliability of the correlation analysis. The parameter influence coefficient calculated in the above manner can comprehensively reflect the strength of the parameter influence between two monitoring points, providing a quantitative basis for subsequent construction of parameter correlation networks.

[0094] Example 4: See Figure 4 When determining the parameter influence coefficient between the first and second monitoring points based on correlation, it is necessary to first collect historical operating data from each monitoring point in the set of monitoring points to be analyzed. This data covers parameter records from multiple past operating cycles, including efficiency-dominant parameters such as power and flow rate, secondary efficiency parameters such as medium temperature, and environmental interference parameters such as ambient humidity. During data collection, it is necessary to cover the pump's operation under different loads and environmental conditions, such as full-load operation on weekdays, low-load operation on weekends, and operating data under special environments such as high temperature and humidity. After collection, the data is processed to remove jump values ​​caused by temporary sensor malfunctions and abnormal interruptions caused by external factors such as sudden power outages, ensuring that the historical operating data accurately reflects the normal operating status of the pump.

[0095] The variance of the collected historical operational data is then determined. Variance reflects the degree of dispersion of a parameter within a historical period. By analyzing the variance, the stability of different parameters can be understood. For example, a small variance in the variance of a flow monitoring point indicates that its value has changed relatively smoothly throughout history; while a large variance in the variance of an environmental temperature monitoring point indicates that it is more significantly affected by the external environment. Multiple correlation levels are then defined based on the magnitude of the variance, with different correlation levels corresponding to different ranges of parameter influence. For instance, the range with the smallest variance corresponds to an extremely strong correlation between parameters, the range with slightly larger variance corresponds to a strong correlation, and so on, until the range with the largest variance corresponds to an extremely weak correlation.

[0096] After determining the target correlation level of the calculated first influence weight and the correlation of historical data, the preset coefficient value corresponding to this level is used as the parameter influence coefficient between the first monitoring point and the second monitoring point. For example, if the historical data correlation between the power monitoring point (efficiency dominant parameter) and the flow monitoring point (efficiency dominant parameter) is at a very strong correlation level, and the preset coefficient value corresponding to this level is 0.9, then the parameter influence coefficient between the two is 0.9; if the correlation between the ambient temperature monitoring point (environmental interference parameter) and the medium temperature monitoring point (efficiency secondary parameter) is at a weak correlation level, and the corresponding preset coefficient value is 0.3, then its parameter influence coefficient is 0.3.

[0097] When processing the data from each monitoring point in the set of monitoring points to be analyzed according to priority, the real-time parameter data of each monitoring point in the priority order is extracted first. Assuming the priority order is power monitoring point, flow monitoring point, ambient temperature monitoring point, and medium temperature monitoring point, the real-time data of these monitoring points in the current period and the most recent few periods are extracted sequentially. The extraction of real-time data must ensure timeliness, typically collected every few minutes to ensure timely reflection of the latest parameter changes. After extraction, a time-series feature sequence is constructed based on these real-time parameter data. Each feature in the sequence corresponds one-to-one with a monitoring point in the priority order. For example, the first feature of the sequence is power data, the second is flow data, and the subsequent features are ambient temperature and medium temperature data, with each feature containing corresponding time information, forming a sequence arranged in chronological order.

[0098] The time-series feature sequences are input into an adaptive threshold model, which generates anomaly warning thresholds for each monitoring point based on the data variation patterns in the sequence. The model analyzes the changing trends of parameters in the time-series feature sequences. For example, power parameters rise rapidly during the pump startup phase and then remain relatively stable during the stable operation phase. Based on this trend, the model sets dynamic thresholds for the startup phase and static thresholds for the stable phase for the power monitoring points. For parameters significantly affected by the environment, such as ambient temperature, the model adjusts the thresholds according to seasonal variations, with warning thresholds higher in summer than in winter, to adapt to environmental differences across seasons.

[0099] When monitoring corresponding monitoring points according to the generated abnormal warning thresholds, the system compares the current parameters of the monitoring points with the warning thresholds in real time. When the parameters of a monitoring point exceed the threshold range, the system will trigger an abnormal warning. For example, if the real-time data of a power monitoring point suddenly exceeds the upper limit of the warning threshold, the system will immediately issue a warning signal; if the value of a flow monitoring point remains below the lower limit of the warning threshold, the system will also trigger a warning after several cycles of confirmation. Warning information will be transmitted in various ways, such as displaying a red alarm icon on the monitoring center screen and sending prompt messages to the terminal devices of maintenance personnel, so that relevant personnel can be informed in a timely manner and take corresponding measures.

[0100] Throughout the process, the adaptive threshold model continuously learns new operating data. When the water pump enters a new operating phase, such as long-term low-load operation, the model will gradually adjust the warning threshold to match the new operating state, ensuring the accuracy and applicability of the abnormal warning.

[0101] Example 5: After setting the abnormal warning threshold, it is necessary to continuously monitor the parameter change rate of each monitoring point in the set of monitoring points to be analyzed. The parameter change rate reflects the magnitude of parameter change at each monitoring point within adjacent periods. By tracking this indicator in real time, potential changes in the pump's operating status can be captured promptly. The monitoring process covers all monitoring points to be analyzed, including efficiency-dominant acquisition points such as power sensors and flow sensors, efficiency-secondary acquisition points such as medium temperature sensors, and environmental interference monitoring points such as ambient humidity sensors. The monitoring frequency is consistent with the parameter acquisition cycle, for example, calculating the parameter change rate every 30 minutes to ensure timely detection of abnormal parameter fluctuations.

[0102] When calculating the rate of change of parameters, it is necessary to obtain the parameter values ​​for the current period and the previous period. The change amplitude is obtained by the ratio of the difference between the two values ​​to the parameter value of the previous period. For example, if the parameter value of a power monitoring point was 100kW in the previous period and the parameter value is 105kW in the current period, the corresponding rate of change is 5%. If the parameter value drops to 98kW in the next period, the rate of change will be -6.67%. During the calculation process, it is important to ensure the consistency of the parameter units. For parameters with different units, such as flow rate and head, the rate of change should be expressed as a percentage for consistent comparison and analysis.

[0103] When the rate of change of a parameter exceeds the preset fluctuation threshold for several consecutive cycles, it indicates that the operating state of the water pump may have changed significantly. In this case, it is necessary to re-execute the step of determining the core path of the parameter correlation network. The number of consecutive cycles is set according to the operating stability of the water pump. For example, for a relatively stable industrial water pump, it can be set to 5 cycles; for a small water pump that frequently starts and stops, it can be set to 3 cycles. The preset fluctuation threshold varies depending on the parameter type. For example, the fluctuation threshold for power parameters is set relatively large, while the fluctuation threshold for ambient temperature parameters is set relatively small, to accommodate the inherent characteristics of different parameters.

[0104] When redefining the core path, it is necessary to recalculate the parameter influence coefficients between each monitoring point based on the latest parameter data and construct a new parameter association network. For example, if the rate of change of ambient temperature continuously exceeds the fluctuation threshold, it may lead to an increase in the influence coefficient between the ambient temperature monitoring point and the medium temperature monitoring point. The ambient temperature monitoring point, which was originally in a secondary position, may form strong associations with more nodes in the new network. By analyzing the connection strength and weight distribution of nodes in the new network, the core path with the greatest impact on the current operating status is identified. The monitoring points on the core path are usually those nodes with large parameter change rates and significant impacts on other parameters.

[0105] The priority ranking is adjusted based on the new core path, raising the priority of monitoring points along the core path and lowering the priority of monitoring points farther away. For example, if a traffic monitoring point is found to be a critical node on the core path after reanalysis, its priority will be adjusted from third to first; while the priority of environmental humidity monitoring points, which were originally high but whose impact has now weakened, may be lowered accordingly. The adjusted priority ranking will affect the order of subsequent data processing, ensuring that data from high-priority monitoring points are analyzed and processed first.

[0106] After priority ranking adjustments, the adaptive threshold model regenerates anomaly warning thresholds based on the new temporal feature sequences, allowing the thresholds to adapt to the changed operating conditions. For example, if the priority of a flow monitoring point is increased, the model will pay more attention to subtle changes in its parameters, generating a more sensitive warning threshold; for monitoring points with decreased priority, the threshold may be appropriately relaxed to reduce unnecessary warnings. Through this dynamic adjustment mechanism, the anomaly warning system can be ensured to always remain synchronized with the actual operating status of the pumps, responding promptly to various changes during operation.

[0107] Throughout the adjustment process, the system records the time periods when parameter change rates exceed limits, changes in the core path, and the specific details of priority adjustments, creating an adjustment log. These logs can be used to trace the changes in the pump's operating status, providing reference information for subsequent maintenance and optimization. Simultaneously, the system continues to monitor the adjusted parameter change rates; if consecutive exceedances occur again, the above adjustment process will be repeated, forming a continuous dynamic adaptation mechanism.

[0108] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0109] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for adaptively setting thresholds for monitoring and issuing early warnings of abnormalities in water pump operation, characterized in that, The method is applied to a water pump monitoring system, which has multiple operating parameter acquisition points pre-deployed. The method includes: Collect various operating parameters of the currently running target water pump within a preset period, and determine the real-time operating status of the target water pump based on the operating parameters; Based on the real-time operating status, the operating parameter collection points are divided into efficiency-dominant collection points and efficiency-secondary collection points. Environmental interference monitoring points are determined from the auxiliary monitoring points other than the operating parameter collection points. The efficiency-dominant collection points, efficiency-secondary collection points, and environmental interference monitoring points constitute the set of monitoring points to be analyzed. Determine the parameter influence coefficient between any two related monitoring points in the set of monitoring points to be analyzed, and construct the parameter association network of the set of monitoring points to be analyzed, using the monitoring points in the set of monitoring points to be analyzed as nodes and the parameter influence coefficient as the weight between the connecting nodes. Determine the core path of the parameter association network, and based on the core path, determine the priority order of each monitoring point in the set of monitoring points to be analyzed; The data of each monitoring point in the set of monitoring points to be analyzed are processed sequentially according to the priority order, so as to achieve adaptive setting of the abnormal early warning threshold.

2. The method for adaptively setting the threshold for monitoring and issuing early warning of abnormalities in water pump operation according to claim 1, characterized in that, Determining the real-time operating status of the target water pump based on the operating parameters includes: Based on the operating parameters, calculate the efficiency benchmark values ​​of the target water pump over multiple historical periods; Based on the fluctuation range of the operating parameters corresponding to each efficiency benchmark value, the calculated efficiency benchmark values ​​are grouped to form one or more efficiency state groups. A target efficiency state group that matches the current operating parameters of the target water pump is determined from the efficiency state group, and the state identifier corresponding to the target efficiency state group is used as the real-time operating state of the target water pump.

3. The method for adaptively setting the threshold for monitoring and issuing early warning of abnormalities in water pump operation according to claim 2, characterized in that, Based on the fluctuation range of the operating parameters corresponding to each efficiency benchmark value, the calculated efficiency benchmark values ​​are grouped as follows: Obtain the parameter fluctuation range preset in the water pump monitoring system; For any calculated efficiency benchmark value, identify the parameter fluctuation range to which the fluctuation range of the operating parameters corresponding to the efficiency benchmark value belongs, and classify the efficiency benchmark value into the efficiency state group corresponding to the identified parameter fluctuation range.

4. The method for adaptively setting the threshold for monitoring and issuing early warning of abnormalities in water pump operation according to claim 1, characterized in that, Based on the real-time operating status, the operating parameter collection points are divided into efficiency-dominant collection points and efficiency-secondary collection points, including: Determine the key and non-key impact parameters corresponding to the real-time operating status; A first monitoring range for the critical impact parameter is determined at the operating parameter acquisition points, and a second monitoring range for the non-critical impact parameter is determined at the operating parameter acquisition points. The operating parameter collection points covered by the first monitoring range are determined as efficiency-dominant collection points, and the operating parameter collection points covered by the second monitoring range are determined as efficiency-secondary collection points.

5. The method for adaptively setting the threshold for monitoring and issuing early warning of abnormalities in water pump operation according to claim 1, characterized in that, Environmental interference monitoring points are identified among the auxiliary monitoring points other than the operational parameter acquisition points, including: A first influence radius is set for the efficiency-dominant collection point among the operating parameter collection points, and a second influence radius is set for the efficiency-secondary collection point among the operating parameter collection points; For auxiliary monitoring points other than the operating parameter collection points, if the distance to the efficiency-dominant collection point is within the first influence radius, then the auxiliary monitoring point is selected as the environmental interference monitoring point; if the distance to the efficiency-secondary collection point is within the second influence radius, then the auxiliary monitoring point is selected as the environmental interference monitoring point.

6. The method for adaptively setting the threshold for monitoring and issuing early warning of abnormalities in water pump operation according to claim 1, characterized in that, Determining the parameter influence coefficient between any two related monitoring points in the set of monitoring points to be analyzed includes: For the first and second monitoring points associated in the set of monitoring points to be analyzed, the parameter types of the first and second monitoring points are identified respectively. The parameter types include one of the following: efficiency dominant parameter, efficiency secondary parameter, and environmental disturbance parameter. Based on the identified parameter types, determine the first influence weight of the first monitoring point and the second influence weight of the second monitoring point; Calculate the correlation between the first influence weight and the historical data of the second monitoring point, and determine the parameter influence coefficient between the first monitoring point and the second monitoring point based on the correlation.

7. The method for adaptively setting the threshold for monitoring and issuing early warning of abnormalities in water pump operation according to claim 6, characterized in that, Determining the parameter influence coefficient between the first monitoring point and the second monitoring point based on the correlation includes: Collect historical operating data of each monitoring point in the set of monitoring points to be analyzed, and determine the variance of the collected historical operating data; Based on the volatility variance, multiple correlation levels are defined, and the target correlation level of the calculated first influence weight and the correlation of historical data is determined. The preset coefficient value corresponding to the target correlation level is determined as the parameter influence coefficient between the first monitoring point and the second monitoring point.

8. The method for adaptively setting the threshold for monitoring and issuing early warning of abnormalities in water pump operation according to claim 1, characterized in that, The data of each monitoring point in the set of monitoring points to be analyzed are processed sequentially according to the priority order, so as to achieve adaptive setting of the abnormal early warning threshold, including: Real-time parameter data of each monitoring point in the priority ranking is extracted, and a time-series feature sequence is constructed based on the extracted real-time parameter data. The arrangement order of each feature in the time-series feature sequence is consistent with the priority ranking. The time-series feature sequence is input into an adaptive threshold model to generate an anomaly warning threshold for each monitoring point. Monitoring is performed on the corresponding monitoring points according to the generated abnormal warning thresholds in order to achieve abnormal warning.

9. The method for adaptively setting the threshold for monitoring and issuing early warning of abnormalities in water pump operation according to claim 1, characterized in that, The method further includes: After the abnormal warning threshold is set, the parameter change rate of each monitoring point in the set of monitoring points to be analyzed is continuously monitored. When the rate of change of the parameter exceeds the preset fluctuation threshold for multiple consecutive periods, the step of determining the core path of the parameter association network is re-executed, and the priority sorting is adjusted according to the new core path.

10. The method for adaptively setting the threshold for monitoring and issuing early warning of abnormalities in water pump operation according to claim 4, characterized in that, The key and non-key impact parameters corresponding to the real-time operating status include: Obtain the design parameter table of the target water pump, which includes the rated flow rate, rated head, and rated power. The actual operating parameters in the real-time operating status are compared with the rated parameters in the design parameter table, and the percentage of parameter deviation is calculated. Operating parameters whose parameter deviation percentage is greater than a preset deviation threshold are identified as critical influencing parameters, while operating parameters whose parameter deviation percentage is less than or equal to the preset deviation threshold are identified as non-critical influencing parameters.

Images