High-pressure water pump motor insulation characteristic on-line monitoring method

Abstract

The invention discloses a high-pressure water pump motor insulation characteristic on-line monitoring method, and relates to the field of high-pressure water pumps, and the method comprises the steps: constructing an insulation prediction model, continuously extracting and analyzing an insulation aging data set in a database, carrying out the calculation, and outputting an insulation basic parameter L1, an insulation characteristic parameter L2 and an external influence factor L3 in real time; and the insulation aging index Lhzs is obtained through comprehensive calculation. And then the insulation aging index Lhzs is input into an adaptive learning algorithm, so that the model can automatically adjust and optimize own parameters, the prediction accuracy is improved, and an update coefficient Gi is obtained. And calculating to obtain a predicted insulation aging coefficient Jylh (t) by using the updated model coefficient Ginew and combining with an insulation aging index Lhzs output in real time, predicting the insulation aging of the high-pressure water pump motor, and identifying a potential aging problem in advance, thereby reducing the frequency of non-planned shutdown, optimizing a maintenance strategy, and ensuring that equipment operates in an optimal state.

Description

Technical Field

[0001] The invention relates to the technical field of high-pressure water pumps, and in particular to an online monitoring method for insulation characteristics of a high-pressure water pump motor. Background Art

[0002] The online monitoring method of the insulation characteristics of high-voltage water pump motors belongs to the large field of power equipment status monitoring and intelligent maintenance. In this field, the health of motor insulation is one of the key factors to ensure the reliable operation of equipment, especially in high-voltage equipment, where insulation failure may cause serious equipment damage and safety hazards. Specifically, the online monitoring method of the insulation characteristics of high-voltage water pump motors is an online monitoring technology for the aging of insulation materials in water pump motors. It aims to collect the insulation data of the motor in real time, evaluate its health status and predict potential aging risks, thereby preventing motor failures.

[0003] At present, because traditional insulation testing relies on shutdown inspections and static model analysis at fixed time intervals, the insulation status of the motor during the non-test period cannot be monitored in real time. This situation can easily lead to potential insulation problems being ignored or delayed in discovery, resulting in unforeseen insulation failures. In addition, static models cannot adapt to changes in equipment operating conditions, which may lead to large deviations in the prediction results of aging status. When the degree of aging of the insulation material exceeds expectations and is not handled in time, it may cause motor insulation breakdown, resulting in equipment shutdown, and even equipment damage and production stagnation, seriously affecting the company's production efficiency and safety. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides an online monitoring method for the insulation characteristics of a high-pressure water pump motor, which solves the problems mentioned in the background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for online monitoring of insulation characteristics of a high-pressure water pump motor comprises the following steps: S1. Install a detection device on the high-pressure water pump motor to collect insulation data of the high-pressure water pump motor in real time, then connect the detection device to the central processing module, and then transmit the collected insulation data to the central processing module in real time; S2, the central processing module receives the collected insulation data in real time, pre-processes the insulation data, obtains the insulation aging data group, and then constructs a database to store the insulation aging data group; S3, construct an insulation prediction algorithm model, extract the insulation aging data group in the database in real time by setting a data extraction interface, input it into the relevant algorithm formula, output the insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 in real time, and then perform comprehensive calculation on the output insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 to obtain the insulation aging index Lhzs; S4, and construct an adaptive learning algorithm, use the comprehensively obtained insulation aging index Lhzs, input it into the adaptive learning algorithm, calculate the output model update coefficient Gi, and incrementally update the insulation aging algorithm model; S5. Input the model update coefficient Gi and the insulation aging index Lhzs. In the updated insulation prediction algorithm model, output the predicted insulation aging coefficient Jylh (t), set the first aging threshold F1 and the second aging threshold F2, conduct comparative evaluation, and then generate an early warning level based on the evaluation results, and generate relevant control mechanisms.

[0006] Preferably, said S1 includes S11 and S12; S11. Install a detection device on the high-pressure water pump motor to collect insulation data of the high-pressure water pump motor in real time; The detection device includes a JY-2500 online detection device for insulation characteristics of high-voltage electrical equipment, a dielectric loss detector, a partial discharge detection device, an electric field detector and a gas sensor; The insulation data includes first insulation characteristic data and second insulation characteristic data; The first insulation characteristic data includes insulation resistance Rin, leakage current Ile, ambient temperature Ten, operating time tru and humidity Hen; The second insulation characteristic data includes the AC voltage Va, the current I, the phase difference between the AC voltage Va and the current I , the total number of discharge events Npd, the instantaneous voltage V(t) and the gas current signal Ise; S12, by constructing a central processing module, the detection device and the central processing module are remotely connected through the TCP protocol, and the collected insulation data is transmitted to the central processing module in real time.

[0007] Preferably, S2 includes S21, S22 and S23; S21, the central processing module receives the transmitted insulation data in real time, and preprocesses the first insulation characteristic data in the insulation data, wherein the preprocessing method includes filtering outliers, denoising and normalizing, to obtain a first insulation characteristic data set; S22, performing feature extraction on the second insulation characteristic data in the insulation data to obtain an aging feature set; The aging feature set includes dielectric loss tangent value tanJ of the insulating material, micro partial discharge frequency Fpd, electrical stress accumulation value Acc and environmental hydrogen content H2; The dielectric loss tangent value tanJ of the insulating material is obtained by the AC voltage Va, the current I and the phase difference between the AC voltage Va and the current I in the second insulation characteristic data. , perform associated calculations to obtain, the calculation formula is: ; The micro partial discharge frequency Fpd is obtained by calculating the total number of discharge events Npd in the second insulation characteristic data, and the calculation formula is: , where T represents the total length of the time window; The electrical stress cumulative value Acc is obtained by calculating the instantaneous voltage V(t) in the second insulation characteristic data, and the calculation formula is: , where is the permittivity in vacuum, represents the relative dielectric constant of the material, d represents the thickness of the insulating material, dt represents the time differential, and J represents the upper limit of the integral of the cumulative time; The ambient hydrogen content H2 is calculated and obtained through the gas current signal Ise in the second insulation characteristic data, and the calculation formula is: , where k represents the sensitivity coefficient of the gas sensor; Then the first insulation characteristic data set and the aging characteristic set are aggregated to obtain an insulation aging data set; S23, by constructing a relational database and setting an input port and an output port, the aging data group obtained after preprocessing is input into the relational database through the input port for storage, and is used to extract the insulation aging data group through the output port.

[0008] Preferably, said S3 includes S31, S32, S33 and S34; S31, used to construct an insulation prediction algorithm model, and set the insulation basic algorithm formula, insulation characteristic algorithm formula and external influence algorithm formula, and then extract the insulation aging data group in real time through the output port of the relational database and input it into the insulation prediction algorithm model, and output the insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 respectively; The insulation basic parameter L1 is obtained by calculating the insulation resistance Rin, the leakage current Ile and the ambient temperature Ten in the insulation aging data group into the insulation basic algorithm formula; The insulation basic parameter L1 is calculated and obtained by the following insulation basic algorithm formula; ; In the formula, log represents the logarithmic function, Indicates the relationship between insulation resistance Rin and leakage current Ile, represents the comprehensive nonlinear effect of leakage current Ile and ambient temperature Ten on aging, Represents the reverse contribution of insulation resistance Rin to aging.

[0009] Preferably, S32, the insulation characteristic parameter L2 is obtained by calculating by inputting the dielectric loss tangent value tanJ and the micro partial discharge frequency Fpd of the insulation material in the insulation aging data group into the insulation characteristic algorithm formula; The insulation characteristic parameter L2 is calculated and obtained by the following insulation characteristic algorithm formula; ; In the formula, exp represents the exponential function, It represents the nonlinear change of dielectric loss tangent tanJ of insulating material under the influence of partial discharge frequency Fpd. The exponential relationship represents the complex influence of the product of the dielectric loss tangent value tanJ of the insulating material and the micro partial discharge frequency Fpd on the aging index.

[0010] Preferably, S33, the external influence factor L3 is obtained by inputting the ambient temperature Ten, the electrical stress accumulation value Acc and the ambient hydrogen content H2 in the insulation aging data group into the external influence algorithm formula for calculation; The external impact factor L3 is calculated and obtained by the following external impact algorithm formula; .

[0011] Preferably, S34, by constructing a comprehensive algorithm model, the output insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 are input into the comprehensive algorithm model, and the insulation aging index Lhzs is comprehensively calculated and output; The insulation aging index Lhzs is calculated by the following algorithm formula: ; In the formula, is the error term, represents the first initial model coefficient, represents the second initial model coefficient, represents the third initial model coefficient.

[0012] Preferably, S4, using an adaptive learning algorithm and utilizing the insulation aging index Lhzs obtained in real time, incrementally updating the comprehensive algorithm model to obtain a model update coefficient Gi; The model update coefficient Gi is calculated and obtained by the following adaptive learning algorithm; ; In the formula, Gi new represents the updated model coefficient, Gi old The model coefficients before updating, represents the learning rate, n represents the total number of samples, j represents the sample, and w j represents the weight of sample j, Lhzs new , j represents the insulation aging index Lhzs, Lhzs output by the updated comprehensive algorithm model of sample j old , j represents the insulation aging index Lhzs output by the comprehensive algorithm model of sample j before updating.

[0013] Preferably, said S5 includes S51 and S52; S51, using the updated model coefficient Gi new , combined with the real-time output insulation aging index Lhzs, calculate and obtain the predicted insulation aging coefficient Jylh (t) to predict the insulation aging of the high-pressure water pump motor; The predicted insulation aging coefficient Jylh (t) is calculated and obtained by the following algorithm formula; ; In the formula, represents the time weight coefficient, m represents the number of time points of historical data, T represents the time window size, Lhzs k Indicates the insulation aging index Lhzs obtained at time point k, x i Represents the input feature values ​​at the i-th time point, which usually include the electrical properties of the insulating material and external environmental factors.

[0014] Preferably, S52, the user presets the first aging threshold F1 and the second aging threshold F2, and then extracts the predicted insulation aging coefficient Jylh (t), performs comparative evaluation to predict the state of insulation aging of the high-pressure water pump motor, and generates an early warning mechanism, and the specific evaluation content is as follows; When the predicted insulation aging coefficient Jylh (t) is greater than the first aging threshold F1, it indicates that the insulation characteristics of the high-pressure water pump motor are abnormally aged. At this time, a first-level warning message is generated, a shutdown command is sent immediately, and an immediate maintenance and replacement is prompted; When the second aging threshold F2 ≤ predicted insulation aging coefficient Jylh (t) ≤ first aging threshold F1, it indicates that the insulation characteristics of the high-pressure water pump motor are abnormally aged. At this time, a second-level warning message is generated, indicating that maintenance is required; When the predicted insulation aging coefficient Jylh(t) is less than the second aging threshold F2, it indicates that the aging of the insulation characteristics of the high-pressure water pump motor is normal, and there is no need to generate an early warning and continue monitoring.

[0015] The present invention provides an online monitoring method for the insulation characteristics of a high-pressure water pump motor. It has the following beneficial effects: (1) By installing high-precision detection devices on the high-pressure water pump motor, such as the JY-2500 high-voltage electrical equipment insulation characteristics online detection device, dielectric loss detector, partial discharge detection equipment, etc., the first insulation characteristics data and the second insulation characteristics data can be collected in real time. These data are remotely transmitted to the central processing module via the TCP protocol, and after pre-processing, they form an accurate insulation aging data group. This can continuously monitor the insulation status during the operation of the motor, timely discover and record minor changes in insulation performance, effectively prevent potential risks caused by insulation aging, and improve the protection capability and safe operation level of the equipment.

[0016] (2) This method constructs an insulation prediction model, continuously extracts and analyzes the insulation aging data set in the database, performs calculations, and outputs the insulation basic parameters L1, insulation characteristic parameters L2, and external influencing factors L3 in real time, and obtains the insulation aging index Lhzs by comprehensive calculation. Then, by inputting the insulation aging index Lhzs into the adaptive learning algorithm, the model can automatically adjust and optimize its own parameters, improve the accuracy of the prediction, and obtain the updated coefficient Gi. Using the updated model coefficient Gi new , combined with the real-time output insulation aging index Lhzs, the predicted insulation aging coefficient Jylh (t) is calculated to predict the insulation aging of the high-pressure water pump motor, identify potential aging problems in advance, thereby reducing the frequency of unplanned downtime, optimizing maintenance strategies, and ensuring that the equipment operates in the best condition.

[0017] (3) By setting the first aging threshold F1 and the second aging threshold F2, this method can accurately compare and evaluate the predicted insulation aging coefficient Jylh(t) calculated in real time and generate the corresponding warning level. When the insulation state of the motor is abnormal, it can automatically trigger the first or second level warning and provide clear maintenance suggestions. For example, when Jylh(t) exceeds the first aging threshold F1, a shutdown command will be sent immediately and a maintenance prompt will be given; when Jylh(t) is between the first aging threshold F1 and the second aging threshold F2, a second level warning will be generated, indicating that detailed inspection and repair are required. This intelligent multi-level warning mechanism can significantly improve the response speed of the operation and maintenance team and ensure that maintenance work is completed before problems occur, thereby extending the service life of the equipment, reducing maintenance costs, and comprehensively improving the level of intelligent equipment management. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the steps of the online monitoring method of the insulation characteristics of the high-pressure water pump motor of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Example 1: The present invention provides an online monitoring method for the insulation characteristics of a high-pressure water pump motor. Figure 1 , including the following steps: S1. Install a detection device on the high-pressure water pump motor to collect insulation data of the high-pressure water pump motor in real time, then connect the detection device to the central processing module, and then transmit the collected insulation data to the central processing module in real time; S2, the central processing module receives the collected insulation data in real time, pre-processes the insulation data, obtains the insulation aging data group, and then constructs a database to store the insulation aging data group; S3, construct an insulation prediction algorithm model, extract the insulation aging data group in the database in real time by setting a data extraction interface, input it into the relevant algorithm formula, output the insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 in real time, and then perform comprehensive calculation on the output insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 to obtain the insulation aging index Lhzs; S4, and construct an adaptive learning algorithm, use the comprehensively obtained insulation aging index Lhzs, input it into the adaptive learning algorithm, calculate the output model update coefficient Gi, and incrementally update the insulation aging algorithm model; S5. Input the model update coefficient Gi and the insulation aging index Lhzs. In the updated insulation prediction algorithm model, output the predicted insulation aging coefficient Jylh (t), set the first aging threshold F1 and the second aging threshold F2, conduct comparative evaluation, and then generate an early warning level based on the evaluation results, and generate relevant control mechanisms.

[0021] In this embodiment, by implementing the online monitoring method for the insulation characteristics of the high-pressure water pump motor, the limitations of the traditional insulation detection method can be effectively overcome, and comprehensive, real-time and accurate insulation status monitoring can be achieved. First, the real-time collection and transmission of insulation data ensures the continuous tracking of the aging of the insulation material, avoiding the lag problem caused by relying on regular shutdown inspections. Secondly, the pre-processed data is continuously optimized through the adaptive learning algorithm model, making the prediction results more accurate, significantly improving the prediction ability of the aging state, and thus reducing the risk of unplanned downtime. Finally, the multi-level early warning mechanism performs real-time evaluation by setting aging thresholds, and can take maintenance measures in time when the problem first occurs to prevent sudden equipment failures, effectively improving the equipment's operational reliability and the intelligent level of maintenance management. Overall, this method provides an intelligent and forward-looking solution for the operation and maintenance of high-pressure water pump motors, comprehensively improves the efficiency and safety of equipment management, and solves the shortcomings of traditional monitoring methods.

[0022] Example 2: This example is an explanation of Example 1. Figure 1 , specifically: S1 includes S11 and S12; S11. Install a detection device on the high-pressure water pump motor to collect insulation data of the high-pressure water pump motor in real time; The detection devices include JY-2500 online detection device for insulation characteristics of high-voltage electrical equipment, dielectric loss detector, partial discharge detection equipment, electric field detector and gas sensor; The insulation data includes first insulation characteristic data and second insulation characteristic data; The first insulation characteristic data includes insulation resistance Rin, leakage current Ile, ambient temperature Ten, operating time tru and humidity Hen; The second insulation characteristic data includes the AC voltage Va, the current I, the phase difference between the AC voltage Va and the current I , the total number of discharge events Npd, the instantaneous voltage V(t) and the gas current signal Ise; S12, by constructing a central processing module, the detection device and the central processing module are remotely connected through the TCP protocol, and the collected insulation data is transmitted to the central processing module in real time.

[0023] In this embodiment, by installing a variety of high-precision detection devices on the high-pressure water pump motor, such as the JY-2500 high-voltage electrical equipment insulation characteristics online detection device, dielectric loss detector, partial discharge detection equipment, power plant detector and gas sensor, insulation data can be collected in real time. These data are remotely transmitted to the central processing module via the TCP protocol, ensuring real-time processing and analysis of the data. Compared with the traditional method that relies on regular shutdown inspections, this method significantly improves the real-time and comprehensiveness of data collection, and can detect potential problems in insulation early.

[0024] Example 3: This example is an explanation of Example 2. Figure 1 , specifically: S2 includes S21, S22 and S23; S21, the central processing module receives the transmitted insulation data in real time, and preprocesses the first insulation characteristic data in the insulation data, wherein the preprocessing method includes filtering outliers, denoising and normalizing, to obtain a first insulation characteristic data set; S22, performing feature extraction on the second insulation characteristic data in the insulation data to obtain an aging feature set; The aging feature set includes the dielectric loss tangent value tanJ of the insulation material, the micro partial discharge frequency Fpd, the electrical stress accumulation value Acc and the ambient hydrogen content H2; The dielectric loss tangent value tanJ of the insulating material is obtained by the AC voltage Va, the current I and the phase difference between the AC voltage Va and the current I in the second insulation characteristic data. , perform associated calculations to obtain, the calculation formula is: ; The micro partial discharge frequency Fpd is calculated by the total number of discharge events Npd in the second insulation characteristic data, and the calculation formula is: , where T represents the total length of the time window; The electrical stress accumulation value Acc is calculated by the instantaneous voltage V(t) in the second insulation characteristic data. The calculation formula is: , where is the permittivity in vacuum, represents the relative dielectric constant of the material, d represents the thickness of the insulating material, dt represents the time differential, and J represents the upper limit of the integral of the cumulative time; The ambient hydrogen content H2 is calculated and obtained through the gas current signal Ise in the second insulation characteristic data. The calculation formula is: , where k represents the sensitivity coefficient of the gas sensor; Then the first insulation characteristic data set and the aging characteristic set are aggregated to obtain an insulation aging data set; S23, by constructing a relational database and setting an input port and an output port, the aging data group obtained after preprocessing is input into the relational database through the input port for storage, and is used to extract the insulation aging data group through the output port.

[0025] In this embodiment, the central processing module of the method receives and processes the insulation data of the high-pressure water pump motor in real time, and can perform pre-processing such as filtering outliers, denoising and normalization on the first insulation characteristic data to ensure the accuracy and consistency of the data. At the same time, through feature extraction, key aging features are obtained from the second insulation characteristic data, such as the dielectric loss tangent value tanJ of the insulating material, the micro-partial discharge frequency Fpd, the electrical stress accumulation value Acc and the environmental hydrogen content H2. These features are obtained through precise calculation formulas to form an aging feature set, which is aggregated with the first insulation characteristic data set to generate an insulation aging data set. Subsequently, the data set is stored in a constructed relational database for subsequent extraction and analysis. Compared with traditional methods, this method has a significant improvement in the accuracy and efficiency of data processing, can more accurately reflect the aging state of the insulating material, reduce the interference of data noise and errors, thereby improving the reliability and real-time performance of the prediction. This method effectively solves the problem of untimely and inaccurate data processing in traditional monitoring, and provides a solid data foundation for accurately evaluating the insulation state of the motor and formulating maintenance plans.

[0026] Example 4: This example is an explanation of Example 3. Figure 1 , specifically: S3 includes S31, S32, S33 and S34; S31, used to construct an insulation prediction algorithm model, and set the insulation basic algorithm formula, insulation characteristic algorithm formula and external influence algorithm formula, and then extract the insulation aging data group in real time through the output port of the relational database and input it into the insulation prediction algorithm model, and output the insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 respectively; The insulation basic parameter L1 is obtained by calculating the insulation resistance Rin, leakage current Ile and ambient temperature Ten in the insulation aging data group into the insulation basic algorithm formula; The insulation basic parameter L1 is calculated and obtained by the following insulation basic algorithm formula; ; In the formula, log represents the logarithmic function, Indicates the relationship between insulation resistance Rin and leakage current Ile, represents the comprehensive nonlinear effect of leakage current Ile and ambient temperature Ten on aging, Represents the reverse contribution of insulation resistance Rin to aging.

[0027] S32, the insulation characteristic parameter L2 is obtained by calculating the dielectric loss tangent value tanJ and the micro partial discharge frequency Fpd of the insulation material in the insulation aging data group into the insulation characteristic algorithm formula; The insulation characteristic parameter L2 is calculated and obtained by the following insulation characteristic algorithm formula; ; In the formula, exp represents the exponential function, It represents the nonlinear change of dielectric loss tangent tanJ of insulating material under the influence of partial discharge frequency Fpd. The exponential relationship represents the complex influence of the product of the dielectric loss tangent value tanJ of the insulating material and the micro partial discharge frequency Fpd on the aging index.

[0028] S33, the external influence factor L3 is obtained by inputting the ambient temperature Ten, the electrical stress accumulation value Acc and the ambient hydrogen content H2 in the insulation aging data group into the external influence algorithm formula for calculation; The external impact factor L3 is calculated and obtained by the following external impact algorithm formula; .

[0029] S34, by constructing a comprehensive algorithm model, the output insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 are input into the comprehensive algorithm model, and the insulation aging index Lhzs is output by comprehensive calculation; The insulation aging index Lhzs is calculated by the following algorithm formula; ; In the formula, is the error term, represents the first initial model coefficient, represents the second initial model coefficient, represents the third initial model coefficient.

[0030] In this embodiment, the method can extract the insulation aging data group from the relational database in real time by constructing an insulation prediction algorithm model and combining a variety of algorithm formulas, and calculate the insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 respectively. These parameters are calculated by precise algorithm formulas, taking into full account the interrelationship and complex influence of key factors such as insulation resistance Rin, leakage current Ile, ambient temperature Ten, dielectric loss tangent value tanJ, micro-partial discharge frequency Fpd, electrical stress accumulation value Acc and ambient hydrogen content H2. Through the comprehensive algorithm model, these parameters are further integrated to output an accurate insulation aging index Lhzs. Compared with traditional methods, this method can deeply analyze various key factors affecting insulation aging, and provide a more comprehensive and accurate aging state assessment through a multi-dimensional algorithm model. This method significantly improves the accuracy and reliability of insulation aging prediction, effectively solves the problem of single model and inaccurate prediction results in traditional prediction methods, thereby providing a solid basis for formulating a scientific equipment maintenance plan and reducing the risk of equipment failure caused by inaccurate aging assessment.

[0031] Example 5: This example is an explanation of Example 4. Figure 1 ,Specifically: S4, through the adaptive learning algorithm, using the insulation aging index Lhzs obtained in real time, incrementally update the comprehensive algorithm model to obtain the model update coefficient Gi; The model update coefficient Gi is calculated and obtained through the following adaptive learning algorithm; ; In the formula, Gi new represents the updated model coefficient, Gi old The model coefficients before updating, represents the learning rate, n represents the total number of samples, j represents the sample, and w j represents the weight of sample j, Lhzs new , j represents the insulation aging index Lhzs, Lhzs output by the updated comprehensive algorithm model of sample j old , j represents the insulation aging index Lhzs output by the comprehensive algorithm model of sample j before updating.

[0032] In this embodiment, the method can use the insulation aging index Lhzs obtained in real time to incrementally update the comprehensive algorithm model through an adaptive learning algorithm, thereby continuously optimizing and adjusting the accuracy of the model. By calculating the model update coefficient Gi, the model parameters can be dynamically adjusted to better reflect the characteristics and trends of the current data. Specifically, by comparing the insulation aging index Lhzs output by the comprehensive algorithm model before and after the update, the model coefficient is accurately adjusted in combination with the learning rate and sample weight. This adaptive learning capability enables the model to evolve continuously with the input of new data, effectively solving the problem that traditional static models cannot adapt to complex and changing environments. Ultimately, this process significantly improves the accuracy and timeliness of insulation aging status prediction, ensures that the best maintenance recommendations can be continuously provided, and reduces prediction deviations caused by model rigidity, thereby greatly improving the reliability of equipment operation and the level of intelligent management.

[0033] Example 6: This example is an explanation of Example 5. Please refer to Figure 1 , specifically: S5 includes S51 and S52; S51, using the updated model coefficient Gi new , combined with the real-time output insulation aging index Lhzs, calculate and obtain the predicted insulation aging coefficient Jylh (t) to predict the insulation aging of the high-pressure water pump motor; The predicted insulation aging factor Jylh (t) is calculated by the following algorithm formula; ; In the formula, represents the time weight coefficient, m represents the number of time points of historical data, T represents the time window size, Lhzs k Indicates the insulation aging index Lhzs obtained at time point k, x i Represents the input feature values ​​at the i-th time point, which usually include the electrical properties of the insulating material and external environmental factors.

[0034] S52, the user presets the first aging threshold F1 and the second aging threshold F2, and then extracts the predicted insulation aging coefficient Jylh(t), performs comparative evaluation to predict the insulation aging status of the high-pressure water pump motor, and generates an early warning mechanism. The specific evaluation contents are as follows; When the predicted insulation aging coefficient Jylh (t) is greater than the first aging threshold F1, it indicates that the insulation characteristics of the high-pressure water pump motor are abnormally aged. At this time, a first-level warning message is generated, a shutdown command is sent immediately, and an immediate maintenance and replacement is prompted; When the second aging threshold F2 ≤ predicted insulation aging coefficient Jylh (t) ≤ first aging threshold F1, it indicates that the insulation characteristics of the high-pressure water pump motor are abnormally aged. At this time, a second-level warning message is generated, indicating that maintenance is required; When the predicted insulation aging coefficient Jylh(t) is less than the second aging threshold F2, it indicates that the aging of the insulation characteristics of the high-pressure water pump motor is normal, and there is no need to generate an early warning and continue monitoring.

[0035] In this embodiment, the method uses the updated model coefficient Gi new Combined with the real-time output insulation aging index Lhzs, the predicted insulation aging coefficient Jylh(t) can be accurately calculated to predict the insulation aging status of the high-pressure water pump motor. This prediction process not only takes into account historical data and time windows, but also combines the electrical properties of the insulating material and external environmental factors, making the prediction results more comprehensive and accurate. On this basis, by comparing the prediction coefficient Jylh(t) with the preset first aging threshold F1 and second aging threshold F2, an early warning mechanism is automatically generated. When the prediction results show abnormal aging, a first-level early warning message or a second-level early warning message can be quickly issued to prompt the maintenance team to take immediate measures or arrange maintenance. This mechanism based on dynamic prediction and multi-level early warning not only improves the response speed to insulation aging problems, but also greatly reduces sudden equipment failures caused by untimely handling of aging. Compared with traditional methods, this effectively makes up for the shortcomings of early warning lag and inaccurate prediction, and significantly improves the safety, reliability and intelligent management level of equipment operation. Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and alterations may be made to the 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. An online monitoring method for insulation characteristics of a high-pressure water pump motor, characterized in that: The following steps are involved: S1. Install a detection device on the high-pressure water pump motor to collect insulation data of the high-pressure water pump motor in real time, then connect the detection device to the central processing module, and then transmit the collected insulation data to the central processing module in real time; S2, the central processing module receives the collected insulation data in real time, pre-processes the insulation data, obtains the insulation aging data group, and then constructs a database to store the insulation aging data group; S3, construct an insulation prediction algorithm model, extract the insulation aging data group in the database in real time by setting a data extraction interface, input it into the relevant algorithm formula, output the insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 in real time, and then perform comprehensive calculation on the output insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 to obtain the insulation aging index Lhzs; S4, and construct an adaptive learning algorithm, use the comprehensively obtained insulation aging index Lhzs, input it into the adaptive learning algorithm, calculate the output model update coefficient Gi, and incrementally update the insulation aging algorithm model; S5. Input the model update coefficient Gi and the insulation aging index Lhzs. In the updated insulation prediction algorithm model, output the predicted insulation aging coefficient Jylh (t), set the first aging threshold F1 and the second aging threshold F2, conduct comparative evaluation, and then generate an early warning level based on the evaluation results, and generate relevant control mechanisms.

2. The method for online monitoring of insulation characteristics of a high-pressure water pump motor according to claim 1, characterized in that: Said S1 includes S11 and S12; S11. Install a detection device on the high-pressure water pump motor to collect insulation data of the high-pressure water pump motor in real time; The detection device includes an online detection device for insulation characteristics of high-voltage electrical equipment, a dielectric loss detector, a partial discharge detection device, an electric field detector and a gas sensor; The insulation data includes first insulation characteristic data and second insulation characteristic data; The first insulation characteristic data includes insulation resistance Rin, leakage current Ile, ambient temperature Ten, operating time tru and humidity Hen; The second insulation characteristic data includes the AC voltage Va, the current I, the phase difference between the AC voltage Va and the current I , the total number of discharge events Npd, the instantaneous voltage V(t) and the gas current signal Ise; S12, by constructing a central processing module, the detection device and the central processing module are remotely connected through the TCP protocol, and the collected insulation data is transmitted to the central processing module in real time.

3. The method for online monitoring insulation characteristics of a high-pressure water pump motor according to claim 1 is characterized in that: The S2 includes S21, S22 and S23; S21, the central processing module receives the transmitted insulation data in real time, and preprocesses the first insulation characteristic data in the insulation data, wherein the preprocessing method includes filtering outliers, denoising and normalizing, to obtain a first insulation characteristic data set; S22, performing feature extraction on the second insulation characteristic data in the insulation data to obtain an aging feature set; The aging feature set includes dielectric loss tangent value tanJ of the insulating material, micro partial discharge frequency Fpd, electrical stress accumulation value Acc and environmental hydrogen content H2; The dielectric loss tangent value tanJ of the insulating material is obtained by the AC voltage Va, the current I and the phase difference between the AC voltage Va and the current I in the second insulation characteristic data. , perform associated calculations to obtain, the calculation formula is: ; The micro partial discharge frequency Fpd is obtained by calculating the total number of discharge events Npd in the second insulation characteristic data, and the calculation formula is: , where T represents the total length of the time window; The electrical stress cumulative value Acc is obtained by calculating the instantaneous voltage V(t) in the second insulation characteristic data, and the calculation formula is: , where is the permittivity in vacuum, represents the relative dielectric constant of the material, d represents the thickness of the insulating material, dt represents the time differential, and J represents the upper limit of the integral of the cumulative time; The ambient hydrogen content H2 is calculated and obtained through the gas current signal Ise in the second insulation characteristic data, and the calculation formula is: , where k represents the sensitivity coefficient of the gas sensor; Then the first insulation characteristic data set and the aging characteristic set are aggregated to obtain an insulation aging data set; S23, by constructing a relational database and setting an input port and an output port, the aging data group obtained after preprocessing is input into the relational database through the input port for storage, and is used to extract the insulation aging data group through the output port.

4. The method for online monitoring insulation characteristics of a high-pressure water pump motor according to claim 3 is characterized in that: The S3 includes S31, S32, S33 and S34; S31, used to construct an insulation prediction algorithm model, and set the insulation basic algorithm formula, insulation characteristic algorithm formula and external influence algorithm formula, and then extract the insulation aging data group in real time through the output port of the relational database and input it into the insulation prediction algorithm model, and output the insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 respectively; The insulation basic parameter L1 is obtained by calculating the insulation resistance Rin, the leakage current Ile and the ambient temperature Ten in the insulation aging data group into the insulation basic algorithm formula; The insulation basic parameter L1 is calculated and obtained by the following insulation basic algorithm formula; ； In the formula, log represents the logarithmic function, Indicates the relationship between insulation resistance Rin and leakage current Ile, represents the comprehensive nonlinear effect of leakage current Ile and ambient temperature Ten on aging, Represents the reverse contribution of insulation resistance Rin to aging.

5. The method for online monitoring insulation characteristics of a high-pressure water pump motor according to claim 4 is characterized in that: S32, the insulation characteristic parameter L2 is obtained by calculating by inputting the dielectric loss tangent value tanJ and the micro partial discharge frequency Fpd of the insulation material in the insulation aging data group into the insulation characteristic algorithm formula; The insulation characteristic parameter L2 is calculated and obtained by the following insulation characteristic algorithm formula; ； In the formula, exp represents the exponential function, It represents the nonlinear change of dielectric loss tangent tanJ of insulating material under the influence of partial discharge frequency Fpd. The exponential relationship represents the complex influence of the product of the dielectric loss tangent value tanJ of the insulating material and the micro partial discharge frequency Fpd on the aging index.

6. The method for online monitoring of insulation characteristics of a high-pressure water pump motor according to claim 4 is characterized in that: S33, the external influence factor L3 is obtained by inputting the ambient temperature Ten, the electrical stress accumulation value Acc and the ambient hydrogen content H2 in the insulation aging data group into the external influence algorithm formula for calculation; The external impact factor L3 is calculated and obtained by the following external impact algorithm formula; 。 7. The method for online monitoring insulation characteristics of a high-pressure water pump motor according to claim 6 is characterized in that: S34, by constructing a comprehensive algorithm model, the output insulation basic parameter L1, insulation characteristic parameter L2 and external influence factor L3 are input into the comprehensive algorithm model, and the insulation aging index Lhzs is output by comprehensive calculation; The insulation aging index Lhzs is calculated by the following algorithm formula: ； In the formula, is the error term, represents the first initial model coefficient, represents the second initial model coefficient, represents the third initial model coefficient.

8. The method for online monitoring insulation characteristics of a high-pressure water pump motor according to claim 4 is characterized in that: S4, using the adaptive learning algorithm and the insulation aging index Lhzs obtained in real time, incrementally updating the comprehensive algorithm model to obtain the model update coefficient Gi; The model update coefficient Gi is calculated and obtained by the following adaptive learning algorithm; ； In the formula, Gi new represents the updated model coefficient, Gi old The model coefficients before updating, represents the learning rate, n represents the total number of samples, j represents the sample, and w j represents the weight of sample j, Lhzs new , j represents the insulation aging index Lhzs, Lhzs output by the updated comprehensive algorithm model of sample j old , j represents the insulation aging index Lhzs output by the comprehensive algorithm model of sample j before updating.

9. The method for online monitoring insulation characteristics of a high-pressure water pump motor according to claim 8, characterized in that: The S5 includes S51 and S52; S51, using the updated model coefficient Gi new , combined with the real-time output insulation aging index Lhzs, calculate and obtain the predicted insulation aging coefficient Jylh (t) to predict the insulation aging of the high-pressure water pump motor; The predicted insulation aging coefficient Jylh (t) is calculated and obtained by the following algorithm formula; ； In the formula, represents the time weight coefficient, m represents the number of time points of historical data, T represents the time window size, Lhzs k Indicates the insulation aging index Lhzs obtained at time point k, x i Represents the input feature value at the i-th time point.

10. The method for online monitoring insulation characteristics of a high-pressure water pump motor according to claim 9, characterized in that: S52, the user presets the first aging threshold F1 and the second aging threshold F2, and then extracts the predicted insulation aging coefficient Jylh(t), performs comparative evaluation to predict the insulation aging status of the high-pressure water pump motor, and generates an early warning mechanism. The specific evaluation contents are as follows; When the predicted insulation aging coefficient Jylh (t) is greater than the first aging threshold F1, it indicates that the insulation characteristics of the high-pressure water pump motor are abnormally aged. At this time, a first-level warning message is generated, a shutdown command is sent immediately, and an immediate maintenance and replacement is prompted; When the second aging threshold F2 ≤ predicted insulation aging coefficient Jylh (t) ≤ first aging threshold F1, it indicates that the insulation characteristics of the high-pressure water pump motor are abnormally aged. At this time, a second-level warning message is generated, indicating that maintenance is required; When the predicted insulation aging coefficient Jylh(t) is less than the second aging threshold F2, it indicates that the aging of the insulation characteristics of the high-pressure water pump motor is normal, and there is no need to generate an early warning and continue monitoring.

Images