Intelligent energy efficiency estimation method and system for submersible pump

Abstract

The invention discloses an intelligent energy efficiency estimation method and system for a submersible pump, and belongs to the technical field of submersible pump energy efficiency optimization, and the method comprises the steps of submersible pump data collection, submersible pump feature engineering, energy efficiency loss evaluation, conveying medium evaluation and submersible pump energy efficiency estimation. According to the invention, a dual-path comprehensive energy efficiency estimation method combining energy efficiency loss evaluation and conveying medium evaluation is adopted, and two intelligent functions are combined, so that multi-dimensional and whole-process accurate evaluation of the operation state of the equipment is realized, and the accuracy and reliability of energy efficiency estimation are improved; energy consumption loss evaluation is carried out by adopting a flow vibration analysis method based on fuzzy intensity, so that accurate identification and quantitative analysis of various loss characteristics under complex working conditions are realized, the equipment state monitoring sensitivity is improved, and the accuracy and reliability of energy efficiency loss evaluation are enhanced; a multi-factor medium correction coefficient calculation method combined with coupling loss graph neural network prediction is adopted to carry out conveying medium evaluation, and online intelligent perception and dynamic correction of medium characteristics are achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of submersible pump energy efficiency optimization, and specifically refers to an intelligent energy efficiency estimation method and system for submersible pumps. Background Art

[0002] As an indispensable core device in the fields of water treatment, agricultural irrigation, urban water supply, etc., submersible pumps directly affect production costs, energy consumption, and environmental protection. The intelligent energy efficiency estimation of submersible pumps uses data analysis and machine learning technologies to intelligently evaluate the energy efficiency of submersible pumps during actual operation from multiple dimensions. The aim is to evaluate the efficiency state of submersible pumps in real time and accurately, assist relevant personnel in promptly detecting abnormalities in submersible pumps, provide data support for the maintenance and management of submersible pumps, and thereby improve the operating efficiency of submersible pumps.

[0003] However, in the existing intelligent energy efficiency estimation process of submersible pumps, there are technical problems such as the design of existing energy efficiency estimation models often failing to fully consider changes in actual operating conditions, such as fluctuations in medium characteristics and common mechanical component wear, which leads to a large deviation between the estimation results and the actual situation; in the existing energy loss assessment process, there are technical problems that the cavitation effect and mechanical wear characteristics are prone to interfere with each other, and early weak fault signals are difficult to capture, resulting in a single mechanical state monitoring method and insufficient feature extraction, thereby affecting the accuracy and reliability of energy loss assessment; in the existing conveying medium assessment process, there are technical problems that traditional medium characteristic detection methods rely on manual sampling and offline analysis, with poor real-time performance and difficulty in adapting to complex and changeable medium environments. Summary of the Invention

[0004] In view of the above situation, to overcome the defects of the prior art, the present invention provides an intelligent energy efficiency estimation method and system for submersible pumps. In the existing intelligent energy efficiency estimation process of submersible pumps, the design of the existing energy efficiency estimation model often fails to fully consider the changes in actual operating conditions, such as fluctuations in medium characteristics and common mechanical component wear, which leads to a large deviation between the estimation result and the actual situation. This solution creatively adopts a dual-path comprehensive energy efficiency estimation method that combines energy efficiency loss assessment and conveying medium assessment. By combining the two intelligent functions, a multi-physical field coupling correction model is constructed for comprehensive estimation, realizing multi-dimensional and whole-process accurate assessment of the equipment operating state, and improving the accuracy and reliability of energy efficiency estimation; in the existing energy efficiency loss assessment process, there is a problem that the cavitation effect and mechanical wear characteristics are prone to interfere with each other, and early weak fault signals are difficult to capture, resulting in a single mechanical state monitoring method and insufficient feature extraction, thus affecting the accuracy and reliability of energy efficiency loss assessment. This solution creatively adopts a flow vibration analysis method based on fuzzy intensity for energy efficiency loss assessment, realizing accurate identification and quantitative analysis of various loss characteristics under complex working conditions, realizing refined analysis of multi-dimensional operation characteristics, effectively improving the sensitivity of equipment state monitoring, and enhancing the accuracy and reliability of energy efficiency loss assessment; in the existing conveying medium assessment process, there is a problem that the traditional medium characteristic detection method relies on manual sampling and offline analysis, with poor real-time performance and difficulty in adapting to complex and changeable medium environments. This solution creatively adopts a multi-factor medium correction coefficient calculation method that combines the prediction of a coupling loss graph neural network for conveying medium assessment, realizing online intelligent perception and dynamic correction of medium characteristics.

[0005] The technical solution adopted by the present invention is as follows: An intelligent energy efficiency estimation method for submersible pumps provided by the present invention includes the following steps:

[0006] Step S1: Submersible pump data collection;

[0007] Step S2: Submersible pump feature engineering;

[0008] Step S3: Energy efficiency loss assessment;

[0009] Step S4: Conveying medium assessment;

[0010] Step S5: Submersible pump energy efficiency estimation.

[0011] Further, in step S1, the submersible pump data collection is used to collect and preprocess the original data. Specifically, electrical parameters, hydraulic parameters, and medium characteristic parameters of the submersible pump are collected in real time by deploying sensors, obtaining the original operation data of the submersible pump, and performing data preprocessing to obtain the standardized operation data of the submersible pump.

[0012] Further, in step S2, the submersible pump feature engineering is used to extract the operation features of the submersible pump. Specifically, multi-scale wavelet cepstrum joint analysis is adopted for the submersible pump feature engineering to obtain the operation features of the submersible pump, including the following steps:

[0013] Step S21: Wavelet decomposition is used to capture changes at different time scales. Specifically, multi-scale analysis is performed on the operation specification data of the submersible pump through wavelet decomposition to obtain multi-band information;

[0014] Step S22: Cepstrum analysis is used to analyze periodic patterns. Specifically, cepstrum analysis is performed on the multi-band information to extract the resonance peak energy and obtain the cepstrum feature information;

[0015] Step S23: Feature integration. Specifically, the multi-band information and the cepstrum feature information are integrated to obtain the operation features of the submersible pump.

[0016] Further, in step S3, the energy efficiency loss assessment is used to evaluate the hidden energy efficiency loss of the submersible pump. Specifically, based on the operation features of the submersible pump, the flow-induced vibration analysis method based on fuzzy intensity is adopted for the energy efficiency loss assessment to obtain the flow-induced vibration analysis data of the pump body of the submersible pump, including the following steps:

[0017] Step S31: Flow-induced vibration signal extraction. Specifically, the features required for the energy efficiency loss assessment are extracted from the operation features of the submersible pump to obtain the flow-induced vibration energy spectrum diagram features of the pump body, and the flow-induced vibration energy spectrum diagram features of the pump body are used to represent the vibration features caused by the flow of the medium in the pump during the operation of the submersible pump;

[0018] Step S32: Cavitation feature modeling is used to quantify the intensity and influence of the cavitation phenomenon during the operation of the pump body. Specifically, the feature information corresponding to the cavitation feature frequency band is extracted from the flow-induced vibration energy spectrum diagram features of the pump body to obtain the cavitation energy index features, and the cavitation fuzzy intensity function based on the sigmoid function is constructed for cavitation fuzzy intensity modeling to obtain the cavitation feature analysis data;

[0019] Step S33: Wear feature modeling is used to quantify the trend of energy efficiency decline caused by the wear of the moving parts in the pump. Specifically, from the flow-induced vibration energy spectrum diagram features of the pump body, sine modulation enhancement is performed according to the wear frequency, and the enhanced cepstrum index features are extracted, and the wear fuzzy intensity function based on the hyperbolic tangent function is used for wear fuzzy intensity modeling to obtain the wear feature analysis data;

[0020] Step S34: Comprehensive fault risk modeling. Specifically, by constructing a weighted neural network, the cavitation feature analysis data and wear feature analysis data are fused to obtain the fault risk coupling intensity feature data. Then, by constructing a gated recurrent attention network combined with dynamic Bayesian residual correction, energy efficiency loss analysis is performed based on the fault risk coupling intensity feature data to obtain the energy efficiency loss data;

[0021] The gated recurrent attention network combined with dynamic Bayesian residual correction includes a gated recurrent attention network and a dynamic Bayesian residual correction model;

[0022] The gated recurrent attention network is used to generate a preliminary prediction value of the energy efficiency loss;

[0023] The dynamic Bayesian residual correction model is used to dynamically correct the prediction value of the energy efficiency loss according to the prediction error;

[0024] Step S35: Energy efficiency loss modeling. Specifically, through the flow vibration signal extraction, the cavitation feature modeling, the wear feature modeling, and the comprehensive fault risk modeling, the energy efficiency loss assessment model is trained to obtain the energy efficiency loss assessment model;

[0025] Step S36: Pump body flow vibration analysis. Specifically, based on the operation characteristics of the submersible pump, the energy efficiency loss assessment is carried out through the energy efficiency loss assessment model to obtain the submersible pump body flow vibration analysis data, which includes the pump body flow vibration energy spectrum diagram characteristics, cavitation feature analysis data, wear feature analysis data, and energy efficiency loss data.

[0026] Further, in step S4, the transport medium evaluation is used to perceive the physical properties of the transport medium. Specifically, based on the operation characteristics of the submersible pump, a multi-factor medium correction coefficient calculation method combined with the prediction of the coupled loss graph neural network is adopted for the transport medium evaluation to obtain the submersible pump transport medium adaptive analysis data, including the following steps:

[0027] Step S41: Medium characteristic extraction. Specifically, the medium characteristic features are extracted from the submersible pump operation specification data, and the medium characteristic features are used to represent the physical properties of the medium;

[0028] Step S42: Medium map construction. Specifically, based on the medium characteristic features, a medium map is constructed;

[0029] Each node of the medium map represents a medium sample, and the node feature is the medium characteristic feature; the edges are constructed by calculating the similarity between each pair of nodes. When the similarity between two nodes is greater than 0.6, an edge is established between these two nodes, and the edge weight is set to the similarity;

[0030] Step S43: medium type modeling, specifically, combining the physical constraint loss function and the cross entropy loss function to construct a coupling loss function, and constructing a graph neural network based on coupling loss to model the medium type according to the medium map to obtain medium type data;

[0031] Step S44: constructing a multi-factor coupling correction coefficient for correcting the energy efficiency estimation data of the submersible pump, specifically constructing a correction coefficient parameter group based on the medium type data, and calculating the multi-factor coupling correction coefficient based on the medium characteristic and the correction coefficient parameter group;

[0032] Step S45: Comprehensive evaluation of the conveying medium, specifically, performing a comprehensive evaluation of the conveying medium through the medium characteristic extraction, the medium spectrum construction, the medium type modeling and the multi-factor coupling correction coefficient construction to obtain the submersible pump conveying medium adaptive analysis data, the submersible pump conveying medium adaptive analysis data including medium characteristic features, medium spectrum, medium type data and multi-factor coupling correction coefficient.

[0033] Further, in step S5, the submersible pump energy efficiency estimation is used to perform a comprehensive estimation of the submersible pump energy efficiency by combining the results of the energy efficiency loss assessment and the conveying medium assessment. Specifically, a multi-physical field coupling correction model is constructed based on the submersible pump body flow vibration analysis data and the submersible pump conveying medium adaptive analysis data, and a comprehensive calculation of the submersible pump energy efficiency value is performed to obtain the submersible pump energy efficiency value estimation reference data. The calculation formula is:

[0034] ;

[0035] ;

[0036] In the formula, It is the comprehensive calculation result of the energy efficiency value of the submersible pump. It is the standard energy efficiency of submersible pumps. is the energy efficiency loss, is the dielectric loss, is the aging attenuation coefficient, e is the base of the natural logarithm, is the material attenuation constant, which is used to indicate the attenuation rate of the material constituting the submersible pump. pump is the operating time of the submersible pump, K factor is the multi-factor coupling correction coefficient.

[0037] The present invention provides an intelligent energy efficiency estimation system for a submersible pump, comprising: a submersible pump data acquisition module, a submersible pump feature engineering module, an energy efficiency loss evaluation module, a conveying medium evaluation module and a submersible pump energy efficiency estimation module;

[0038] The submersible pump data acquisition module is used for submersible pump data acquisition. Through submersible pump data acquisition, the operation specification data of the submersible pump is obtained, and the operation specification data of the submersible pump is sent to the submersible pump feature engineering module and the conveying medium evaluation module;

[0039] The submersible pump feature engineering module is used for submersible pump feature engineering. Through submersible pump feature engineering, the operation characteristics of the submersible pump are obtained, and the operation characteristics of the submersible pump are sent to the energy efficiency loss evaluation module;

[0040] The energy efficiency loss evaluation module is used for energy efficiency loss evaluation. Through energy efficiency loss evaluation, the flow vibration analysis data of the submersible pump body is obtained, and the flow vibration analysis data of the submersible pump body is sent to the submersible pump energy efficiency estimation module;

[0041] The conveying medium evaluation module is used for conveying medium evaluation. Through conveying medium evaluation, the adaptive analysis data of the submersible pump conveying medium is obtained, and the adaptive analysis data of the submersible pump conveying medium is sent to the submersible pump energy efficiency estimation module;

[0042] The submersible pump energy efficiency estimation module is used for submersible pump energy efficiency estimation. Through submersible pump energy efficiency estimation, the reference data for estimating the energy efficiency value of the submersible pump is obtained.

[0043] The beneficial effects achieved by the present invention using the above solution are as follows:

[0044] (1) Aiming at the technical problem that in the existing intelligent energy efficiency estimation process of submersible pumps, the design of existing energy efficiency estimation models often fails to fully consider the changes in actual operating conditions, such as fluctuations in medium characteristics and common mechanical component wear, which leads to a large deviation between the estimation results and the actual situation. This solution creatively adopts a dual-path comprehensive energy efficiency estimation method combining energy efficiency loss evaluation and conveying medium evaluation. By combining the two intelligent functions, a multi-physical field coupling correction model is constructed for comprehensive estimation, realizing multi-dimensional and whole-process accurate evaluation of the equipment operation state, and improving the accuracy and reliability of energy efficiency estimation;

[0045] (2) Aiming at the technical problem that in the existing energy efficiency loss evaluation process, the cavitation effect and mechanical wear characteristics are prone to interfere with each other, and early weak fault signals are difficult to capture, resulting in a single mechanical state monitoring method and insufficient feature extraction, thus affecting the accuracy and reliability of energy efficiency loss evaluation. This solution creatively adopts a flow vibration analysis method based on fuzzy intensity for energy efficiency loss evaluation, realizing accurate identification and quantitative analysis of various loss characteristics under complex working conditions, realizing refined analysis of multi-dimensional operation characteristics, effectively improving the sensitivity of equipment state monitoring, and enhancing the accuracy and reliability of energy efficiency loss evaluation;

[0046] (3) In view of the technical problems existing in the existing conveying medium evaluation process, where the traditional medium characteristic detection method relies on manual sampling and offline analysis, has poor real-time performance and is difficult to adapt to complex and changeable medium environments, this solution creatively adopts a multi-factor medium correction coefficient calculation method combined with the prediction of a coupling loss graph neural network to evaluate the conveying medium, achieving online intelligent perception and dynamic correction of medium characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic flow chart of an intelligent energy efficiency estimation method for a submersible pump provided by the present invention;

[0048] Figure 2 It is a schematic diagram of an intelligent energy efficiency estimation system for a submersible pump provided by the present invention;

[0049] Figure 3 It is a schematic flow chart of step S3;

[0050] Figure 4 It is a schematic flow chart of step S4.

[0051] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments in the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0053] In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc. indicating the orientation or position relationship are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation to the present invention.

[0054] Embodiment 1, referring to Figure 1 , an intelligent energy efficiency estimation method for a submersible pump provided by the present invention, the method includes the following steps:

[0055] Step S1: Submersible pump data collection;

[0056] Step S2: Submersible pump feature engineering;

[0057] Step S3: Energy efficiency loss assessment;

[0058] Step S4: Transport medium assessment;

[0059] Step S5: Submersible pump energy efficiency estimation;

[0060] By performing the above operations, in the existing intelligent energy efficiency estimation process of submersible pumps, there is a technical problem that the design of existing energy efficiency estimation models often fails to fully consider the changes in actual operating conditions, such as fluctuations in medium characteristics and common mechanical component wear, which leads to a large deviation between the estimation results and the actual situation. This solution creatively adopts a dual-path comprehensive energy efficiency estimation method that combines energy efficiency loss assessment and transport medium assessment. By combining the two intelligent functions, a multi-physical field coupling correction model is constructed for comprehensive estimation, realizing multi-dimensional and whole-process accurate assessment of the equipment operating state, and improving the accuracy and reliability of energy efficiency estimation.

[0061] Example 2, refer to Figure 1 , based on the above example, in step S1, the submersible pump data collection is used to collect and preprocess the original data. Specifically, electrical parameters, hydraulic parameters, and medium characteristic parameters of the submersible pump are collected in real time by deploying sensors, the original operation data of the submersible pump is obtained, and data preprocessing is performed to obtain the standardized operation data of the submersible pump;

[0062] The electrical parameters include three-phase voltage, current, active power, reactive power, power factor, and frequency;

[0063] The hydraulic parameters include flow rate, head, and water pressure;

[0064] The medium characteristic parameters include medium temperature, medium density, medium viscosity, medium gas content, and medium sand content;

[0065] The data preprocessing includes data cleaning, data synchronization alignment, and data standardization.

[0066] Example 3, refer to Figure 1 , based on the above example, in step S2, the submersible pump feature engineering is used to extract the operating features of the submersible pump. Specifically, multi-scale wavelet cepstrum joint analysis is adopted for submersible pump feature engineering to obtain the operating features of the submersible pump, including the following steps:

[0067] Step S21: Wavelet decomposition, which is used to capture changes at different time scales. Specifically, multi-scale analysis of the standardized operation data of the submersible pump is performed through wavelet decomposition to obtain multi-band information;

[0068] Step S22: Cepstrum analysis, which is used to analyze periodic patterns. Specifically, cepstrum analysis is performed on the multi-band information to extract formant energy, and cepstrum feature information is obtained.

[0069] Step S23: Feature integration. Specifically, the multi-band information and the cepstrum feature information are integrated to obtain the operation characteristics of the submersible pump.

[0070] Example 4, refer to Figure 1 and Figure 3 , this example is based on the above example. In step S3, the energy efficiency loss assessment is used to evaluate the hidden energy efficiency loss of the submersible pump. Specifically, based on the operation characteristics of the submersible pump, a flow-induced vibration analysis method based on fuzzy intensity is used for energy efficiency loss assessment to obtain the flow-induced vibration analysis data of the pump body of the submersible pump, including the following steps:

[0071] Step S31: Flow-induced vibration signal extraction. Specifically, the features required for energy efficiency loss assessment are extracted from the operation characteristics of the submersible pump to obtain the flow-induced vibration energy spectrum diagram features of the pump body. The flow-induced vibration energy spectrum diagram features of the pump body are used to represent the vibration characteristics caused by the flow of the medium in the pump during the operation of the submersible pump. The calculation formula is:

[0072] ;

[0073] In the formula, MPA(i,j) is the feature of the flow-induced vibration energy spectrum diagram of the pump body in the i-th frequency band and the j-th time window. i is the frequency band index, j is the time window index, log(·) is the logarithmic function, WPT i,j is the wavelet decomposition result of the i-th frequency band in the j-th time window, is the fusion weight, which is used to control the combination degree of cepstrum analysis and wavelet decomposition. Cep i,j is the cepstrum analysis result of the i-th frequency band in the j-th time window;

[0074] Step S32: Cavitation feature modeling, which is used to quantify the intensity and influence of cavitation phenomena during the operation of the pump body. Specifically, the feature information corresponding to the cavitation feature frequency band is extracted from the flow-induced vibration energy spectrum diagram features of the pump body to obtain the cavitation energy index feature, and a cavitation fuzzy intensity function based on the sigmoid function is constructed for cavitation fuzzy intensity modeling to obtain the cavitation feature analysis data;

[0075] The calculation formula for extracting the feature information corresponding to the cavitation feature frequency band from the flow-induced vibration energy spectrum diagram features of the pump body to obtain the cavitation energy index feature is:

[0076] ;

[0077] In the formula, E cav is the cavitation energy index feature, M is the number of sub-frequency bands of the cavitation feature, w iis the weight of the i-th frequency band, used to adjust the influence degree of different frequency bands. k is the index of the frequency band sample point, and B i is the i-th frequency band range corresponding to the cavitation feature, with a value range of [2 kHz, 5 kHz]. WPT(k) is the wavelet decomposition result at the k-th frequency band sample point;

[0078] The calculation formula of the cavitation fuzzy intensity function based on the sigmoid function is as follows:

[0079] ;

[0080] In the formula, is the cavitation fuzzy intensity function based on the sigmoid function, sigmoid(·) is the sigmoid function, A is the cavitation fuzzy intensity factor, used to control the change speed of the cavitation fuzzy intensity, and B is the cavitation energy offset factor;

[0081] Step S33: Wear feature modeling, used to quantify the trend of energy efficiency decline caused by the wear of moving parts in the pump. Specifically, from the characteristics of the pump body flow vibration energy spectrum, according to the wear frequency, sinusoidal modulation enhancement is carried out, the enhanced cepstrum index feature is extracted, and through the wear fuzzy intensity function based on the hyperbolic tangent function, wear fuzzy intensity modeling is carried out to obtain wear feature analysis data;

[0082] The wear frequencies include bearing fault frequency, impeller fault frequency, and pump shaft fault frequency;

[0083] The calculation formula of the wear fuzzy intensity function based on the hyperbolic tangent function is as follows:

[0084] ;

[0085] In the formula, is the wear fuzzy intensity function based on the hyperbolic tangent function, WI is the enhanced cepstrum index feature, tanh(·) is the hyperbolic tangent function, is the wear fuzzy intensity factor, used to control the change speed of the wear fuzzy intensity;

[0086] Step S34: Fault risk comprehensive modeling. Specifically, by constructing a weighted neural network, the cavitation feature analysis data and the wear feature analysis data are fused in terms of features to obtain fault risk coupling intensity feature data, and by constructing a gated recurrent attention network combined with dynamic Bayesian residual correction, energy efficiency loss analysis is carried out based on the fault risk coupling intensity feature data to obtain energy efficiency loss data;

[0087] The gated recurrent attention network combined with dynamic Bayesian residual correction includes a gated recurrent attention network and a dynamic Bayesian residual correction model;

[0088] The gated recurrent attention network is used to generate a preliminary predicted value of energy efficiency loss;

[0089] The dynamic Bayesian residual correction model is used to dynamically correct the predicted value of energy efficiency loss according to the prediction error;

[0090] The calculation formula for analyzing energy efficiency loss based on the characteristic data of fault risk coupling strength is as follows:

[0091] ;

[0092] In the formula, is the energy efficiency loss at the t-th time step, t is the time step index, z is the fault risk coupling strength characteristic index, N is the number of fault risk coupling strength characteristics, and W z,t is the time-varying weight of the z-th fault risk coupling strength characteristic at the t-th time step, is the z-th fault risk coupling strength characteristic at the t-th time step, is the exponential adjustment coefficient of the z-th fault risk coupling strength characteristic, is the dynamic Bayesian residual correction term at the t-th time step;

[0093] Step S35: Energy efficiency loss modeling, specifically, through the above-mentioned flow vibration signal extraction, cavitation feature modeling, wear feature modeling, and comprehensive fault risk modeling, training an energy efficiency loss evaluation model to obtain an energy efficiency loss evaluation model;

[0094] Step S36: Pump body flow vibration analysis, specifically, based on the operating characteristics of the submersible pump, using the energy efficiency loss evaluation model to evaluate the energy efficiency loss, obtaining submersible pump body flow vibration analysis data, and the submersible pump body flow vibration analysis data includes pump body flow vibration energy spectrum diagram features, cavitation feature analysis data, wear feature analysis data, and energy efficiency loss data;

[0095] By performing the above operations, in view of the technical problems existing in the existing energy efficiency loss evaluation process, where the cavitation effect and mechanical wear characteristics are prone to interfere with each other, and early weak fault signals are difficult to capture, resulting in a single mechanical state monitoring method and insufficient feature extraction, thereby affecting the accuracy and reliability of energy efficiency loss evaluation, this solution creatively uses a flow vibration analysis method based on fuzzy intensity for energy efficiency loss evaluation, realizing accurate identification and quantitative analysis of various loss characteristics under complex working conditions, realizing refined analysis of multi-dimensional operating characteristics, effectively improving the sensitivity of equipment state monitoring, and enhancing the accuracy and reliability of energy efficiency loss evaluation.

[0096] Example Five, refer to Figure 1 and Figure 4, this embodiment is based on the above embodiment. In step S4, the conveying medium evaluation is used to sense the physical properties of the conveying medium. Specifically, according to the operating characteristics of the submersible pump, a multi-factor medium correction coefficient calculation method combined with the prediction of the coupling loss graph neural network is adopted to conduct the conveying medium evaluation, and the adaptive analysis data of the submersible pump conveying medium is obtained, including the following steps:

[0097] Step S41: Extract medium characteristics, specifically extract the medium characteristic features from the submersible pump operation specification data, and the medium characteristic features are used to represent the physical properties of the medium;

[0098] Step S42: Construct a medium map, specifically construct a medium map according to the medium characteristic features;

[0099] Each node of the medium map represents a medium sample, and the node feature is the medium characteristic feature; the edges are constructed by calculating the similarity between each node. When the similarity between two nodes is greater than 0.6, an edge is established between these two nodes, and the edge weight is set to the similarity;

[0100] The calculation formula for the similarity is:

[0101] ;

[0102] In the formula, wig ab is the similarity between the a-th node and the b-th node, a is the first node index, b is the second node index, and the second node index is not equal to the first node index. exp(·) is the exponential function, is the sensitivity factor, which is used to control the influence degree of the physical property difference of the medium on the similarity, is the difference in medium density between the a-th node and the b-th node, is the difference in medium viscosity between the a-th node and the b-th node, is the difference in medium gas content between the a-th node and the b-th node, is the difference in medium sand content between the a-th node and the b-th node;

[0103] Step S43: Model the medium type, specifically combine the physical constraint loss function and the cross-entropy loss function to construct a coupling loss function, and construct a graph neural network based on the coupling loss. According to the medium map, conduct medium type modeling to obtain medium type data;

[0104] The calculation formula for the coupling loss function is:

[0105] ;

[0106] ;

[0107] where L total is the coupling loss function, L cross is the cross-entropy loss function, w phy is the physical constraint weight, L phy is the physical constraint loss function, is the Euclidean distance weight, E m is the medium spectrum, h a is the feature of the a-th node, h b is the feature of the b-th node, ||·||2 is the L2 norm for calculating the Euclidean distance, is the cosine similarity weight, CosSim(·) is the cosine similarity function, S ab is the edge weight between the a-th node and the b-th node;

[0108] Step S44: Construct a multi-factor coupling correction coefficient for correcting the energy efficiency estimation data of the submersible pump. Specifically, according to the medium type data, construct a correction coefficient parameter group, and calculate the multi-factor coupling correction coefficient based on the medium characteristic features and the correction coefficient parameter group. The calculation formula is:

[0109] ;

[0110] where K factor is the multi-factor coupling correction coefficient, is the viscosity weight, tanh(·) is the hyperbolic tangent function, is the medium viscosity, is the standard medium viscosity, and the standard medium specifically refers to clear water, is the sediment concentration weight, C sand is the sediment concentration of the medium, C ref is the reference value of the standard medium sediment concentration, is the sediment concentration influence factor, is the temperature weight, is the difference between the standard medium temperature and the medium temperature;

[0111] Preferably, Table 1 is a comparison parameter table of the correction coefficient parameter group and the medium type. As shown in the table, the medium types include seawater, oily sewage, sediment-laden water, high-temperature and high-pressure liquid, and fresh water;

[0112] Table 1 Comparison parameter table of the correction coefficient parameter group and the medium type

[0113]

[0114] Step S45: Comprehensive evaluation of the conveying medium. Specifically, through the extraction of the medium characteristics, the construction of the medium map, the modeling of the medium type, and the construction of the multi-factor coupling correction coefficient, a comprehensive evaluation of the conveying medium is carried out to obtain the adaptive analysis data of the submersible pump conveying medium. The adaptive analysis data of the submersible pump conveying medium includes medium characteristic features, medium maps, medium type data, and multi-factor coupling correction coefficients;

[0115] By performing the above operations, aiming at the technical problems existing in the existing conveying medium evaluation process, where the traditional medium characteristic detection method relies on manual sampling and offline analysis, with poor real-time performance and difficulty in adapting to complex and changeable medium environments, this solution creatively adopts a multi-factor medium correction coefficient calculation method combined with the prediction of the coupling loss graph neural network to carry out the evaluation of the conveying medium, realizing the online intelligent perception and dynamic correction of the medium characteristics.

[0116] Example Six. Refer to Figure 1 , based on the above example, in step S5, the energy efficiency estimation of the submersible pump is used to comprehensively estimate the energy efficiency of the submersible pump by combining the results of the energy efficiency loss evaluation and the conveying medium evaluation. Specifically, according to the submersible pump pump body flow vibration analysis data and the adaptive analysis data of the submersible pump conveying medium, a multi-physical field coupling correction model is constructed to comprehensively calculate the energy efficiency value of the submersible pump, obtaining the reference data for the energy efficiency value estimation of the submersible pump. The calculation formula is:

[0117] ;

[0118] ;

[0119] In the formula, is the comprehensive calculation result of the submersible pump energy efficiency value, is the standard energy efficiency of the submersible pump, is the energy efficiency loss, is the medium loss, is the aging attenuation coefficient, e is the base of the natural logarithm, is the material attenuation constant, used to represent the attenuation rate of the materials constituting the submersible pump, T pump is the operating duration of the submersible pump, K factor is the multi-factor coupling correction coefficient.

[0120] Example Seven. Refer to Figure 2 , based on the above example, an intelligent energy efficiency estimation system for a submersible pump provided by the present invention includes: a submersible pump data acquisition module, a submersible pump feature engineering module, an energy efficiency loss evaluation module, a conveying medium evaluation module, and a submersible pump energy efficiency estimation module;

[0121] The submersible pump data acquisition module is used for submersible pump data acquisition. Through submersible pump data acquisition, the operation specification data of the submersible pump is obtained, and the operation specification data of the submersible pump is sent to the submersible pump feature engineering module and the conveying medium evaluation module;

[0122] The submersible pump feature engineering module is used for submersible pump feature engineering. Through submersible pump feature engineering, the operation characteristics of the submersible pump are obtained, and the operation characteristics of the submersible pump are sent to the energy efficiency loss evaluation module;

[0123] The energy efficiency loss evaluation module is used for energy efficiency loss evaluation. Through energy efficiency loss evaluation, the flow vibration analysis data of the submersible pump body is obtained, and the flow vibration analysis data of the submersible pump body is sent to the submersible pump energy efficiency estimation module;

[0124] The conveying medium evaluation module is used for conveying medium evaluation. Through conveying medium evaluation, the adaptive analysis data of the submersible pump conveying medium is obtained, and the adaptive analysis data of the submersible pump conveying medium is sent to the submersible pump energy efficiency estimation module;

[0125] The submersible pump energy efficiency estimation module is used for submersible pump energy efficiency estimation. Through submersible pump energy efficiency estimation, the reference data for estimating the energy efficiency value of the submersible pump is obtained.

[0126] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0127] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principle and spirit of the present invention.

[0128] The above describes the present invention and its implementation manners. This description is not restrictive. What is shown in the drawings is only one of the implementation manners of the present invention, and the actual structure is not limited thereto. All in all, if those of ordinary skill in the art are inspired by it and design similar structural manners and embodiments without creative efforts without departing from the purpose of the present invention, they shall fall within the protection scope of the present invention.

Claims

1. A method for intelligent energy efficiency estimation of a submersible pump, characterized in that: The method comprises the following steps: Step S1: collecting submersible pump data to obtain submersible pump operation specification data; Step S2: submersible pump feature engineering, used to extract the submersible pump operation characteristics, specifically using multi-scale wavelet cepstrum joint analysis to perform submersible pump feature engineering to obtain the submersible pump operation characteristics; Step S3: Energy efficiency loss assessment, which is used to assess the hidden energy efficiency loss of the submersible pump. Specifically, based on the operation characteristics of the submersible pump, a flow vibration analysis method based on fuzzy intensity is used to assess the energy efficiency loss, and the flow vibration analysis data of the submersible pump body is obtained; Step S4: transport medium evaluation, which is used to sense the physical characteristics of the transport medium. Specifically, based on the operating characteristics of the submersible pump, a multi-factor medium correction coefficient calculation method combined with the coupling loss graph neural network prediction is used to evaluate the transport medium and obtain the self-adaptive analysis data of the submersible pump transport medium. Step S5: Submersible pump energy efficiency estimation, which is used to perform a comprehensive estimation of the submersible pump energy efficiency by combining the results of the energy efficiency loss assessment and the transport medium assessment.

2. The intelligent energy efficiency estimation method for a submersible pump according to claim 1 is characterized in that: In step S3, the energy efficiency loss assessment includes the following steps: Step S31: extracting flow vibration signals, specifically extracting the features required for energy efficiency loss assessment from the operating features of the submersible pump to obtain pump body flow vibration energy spectrum features, wherein the pump body flow vibration energy spectrum features are used to represent the vibration features caused by the flow of the medium in the pump during the operation of the submersible pump; Step S32: Cavitation feature modeling is used to quantify the intensity and influence of cavitation phenomena during the operation of the pump body. Specifically, the characteristic information of the frequency band corresponding to the cavitation feature is extracted from the pump body flow vibration energy spectrum characteristics to obtain the cavitation energy index characteristics, and the cavitation fuzzy intensity function based on the sigmoid function is constructed to perform cavitation fuzzy intensity modeling to obtain cavitation feature analysis data; Step S33: Wear feature modeling is used to quantify the energy efficiency decline trend caused by the wear of the moving parts in the pump. Specifically, from the pump body flow vibration energy spectrum characteristics, sinusoidal modulation enhancement is performed according to the wear frequency to extract the enhanced cepstrum index characteristics, and wear fuzzy intensity modeling is performed through the wear fuzzy intensity function based on the hyperbolic tangent function to obtain wear feature analysis data; Step S34: comprehensive fault risk modeling, specifically, by constructing a weighted neural network, cavitation feature analysis data and wear feature analysis data are feature-fused to obtain fault risk coupling intensity feature data, and by constructing a gated recurrent attention network combined with dynamic Bayesian residual correction, energy efficiency loss analysis is performed based on the fault risk coupling intensity feature data to obtain energy efficiency loss data; Step S35: energy efficiency loss modeling, specifically, training an energy efficiency loss evaluation model through the flow vibration signal extraction, the cavitation feature modeling, the wear feature modeling and the comprehensive fault risk modeling to obtain an energy efficiency loss evaluation model; Step S36: Pump body flow vibration analysis, specifically, based on the operation characteristics of the submersible pump, energy efficiency loss is evaluated through an energy efficiency loss evaluation model to obtain submersible pump body flow vibration analysis data, wherein the submersible pump body flow vibration analysis data includes pump body flow vibration energy spectrum characteristics, cavitation characteristic analysis data, wear characteristic analysis data and energy efficiency loss data.

3. The intelligent energy efficiency estimation method for a submersible pump according to claim 2 is characterized in that: In step S4, the transport medium evaluation comprises the following steps: Step S41: extracting medium characteristics, specifically extracting medium characteristic features from the submersible pump operation specification data, wherein the medium characteristic features are used to represent the physical characteristics of the medium; Step S42: constructing a medium map, specifically constructing a medium map according to the medium characteristics; Step S43: medium type modeling, specifically, combining the physical constraint loss function and the cross entropy loss function to construct a coupling loss function, and constructing a graph neural network based on coupling loss to model the medium type according to the medium map to obtain medium type data; Step S44: constructing a multi-factor coupling correction coefficient for correcting the energy efficiency estimation data of the submersible pump, specifically constructing a correction coefficient parameter group based on the medium type data, and calculating the multi-factor coupling correction coefficient based on the medium characteristic and the correction coefficient parameter group; Step S45: Comprehensive evaluation of the conveying medium, specifically, performing a comprehensive evaluation of the conveying medium through the medium characteristic extraction, the medium spectrum construction, the medium type modeling and the multi-factor coupling correction coefficient construction to obtain the submersible pump conveying medium adaptive analysis data, the submersible pump conveying medium adaptive analysis data including medium characteristic features, medium spectrum, medium type data and multi-factor coupling correction coefficient.

4. The intelligent energy efficiency estimation method for a submersible pump according to claim 3 is characterized by: In step S34, the gated recurrent attention network combined with dynamic Bayesian residual correction includes a gated recurrent attention network and a dynamic Bayesian residual correction model; The gated recurrent attention network is used to generate a preliminary prediction value of energy efficiency loss; The dynamic Bayesian residual correction model is used to dynamically correct the predicted value of energy efficiency loss according to the prediction error; In step S42, each node of the medium map represents a medium sample, and the node feature is a medium characteristic feature; Edges are constructed by calculating the similarity between each node. When the similarity between two nodes is greater than 0.6, an edge is established between the two nodes and the edge weight is set to the similarity.

5. The intelligent energy efficiency estimation method for a submersible pump according to claim 4 is characterized in that: In step S2, the submersible pump feature engineering is used to extract the submersible pump operation characteristics, specifically using multi-scale wavelet cepstrum joint analysis to perform submersible pump feature engineering to obtain the submersible pump operation characteristics, including the following steps: Step S21: wavelet decomposition is used to capture changes at different time scales, specifically, multi-scale analysis of the submersible pump operation specification data is performed through wavelet decomposition to obtain multi-frequency band information; Step S22: Cepstrum analysis is used to analyze periodic patterns, specifically, performing cepstrum analysis on multi-band information, extracting resonance peak energy, and obtaining cepstrum feature information; Step S23: feature integration, specifically, integrating the multi-band information and the cepstrum feature information to obtain the submersible pump operation characteristics.

6. The intelligent energy efficiency estimation method for a submersible pump according to claim 5 is characterized by: In step S5, the energy efficiency of the submersible pump is estimated by constructing a multi-physical field coupling correction model based on the flow vibration analysis data of the submersible pump body and the adaptive analysis data of the submersible pump conveying medium, and performing a comprehensive calculation of the energy efficiency value of the submersible pump to obtain reference data for the energy efficiency value estimation of the submersible pump. The calculation formula is: ； ； In the formula, It is the comprehensive calculation result of the energy efficiency value of the submersible pump. It is the standard energy efficiency of submersible pumps. is the energy efficiency loss, is the dielectric loss, is the aging attenuation coefficient, e is the base of the natural logarithm, is the material attenuation constant, which is used to indicate the attenuation rate of the material constituting the submersible pump. pump is the operating time of the submersible pump, K factor is the multi-factor coupling correction coefficient.

7. The intelligent energy efficiency estimation method for a submersible pump according to claim 6 is characterized by: In step S1, the submersible pump data acquisition is used to collect raw data and preprocess it. Specifically, sensors are deployed to collect electrical parameters, hydraulic parameters and medium characteristic parameters of the submersible pump in real time, obtain the submersible pump operation raw data, and perform data preprocessing to obtain submersible pump operation specification data.

8. A submersible pump intelligent energy efficiency estimation system, used to implement a submersible pump intelligent energy efficiency estimation method according to any one of claims 1 to 7, characterized in that: It includes a submersible pump data acquisition module, a submersible pump feature engineering module, an energy efficiency loss evaluation module, a conveying medium evaluation module and a submersible pump energy efficiency estimation module.

9. The intelligent energy efficiency estimation system for submersible pumps according to claim 8, characterized in that: The submersible pump data acquisition module is used for submersible pump data acquisition, obtains submersible pump operation specification data through submersible pump data acquisition, and sends the submersible pump operation specification data to the submersible pump feature engineering module and the conveying medium evaluation module; The submersible pump feature engineering module is used for submersible pump feature engineering, obtains submersible pump operation features through submersible pump feature engineering, and sends the submersible pump operation features to the energy efficiency loss assessment module; The energy efficiency loss assessment module is used for energy efficiency loss assessment, and obtains flow vibration analysis data of the submersible pump body through energy efficiency loss assessment, and sends the flow vibration analysis data of the submersible pump body to the submersible pump energy efficiency estimation module; The conveying medium evaluation module is used for conveying medium evaluation, and obtains the submersible pump conveying medium adaptive analysis data through conveying medium evaluation, and sends the submersible pump conveying medium adaptive analysis data to the submersible pump energy efficiency estimation module; The submersible pump energy efficiency estimation module is used for submersible pump energy efficiency estimation, and obtains submersible pump energy efficiency value estimation reference data through submersible pump energy efficiency estimation.

Images