Intelligent detection method and system for particle-resistant magnetic force pump

Abstract

The application discloses a kind of particle-resistant type magnetic force pump intelligent detection method and system, it is related to magnetic force pump detection field, wherein, the method includes: collection operating parameter set;Analysis occurs in multiple operating parameter categories Change and the correlation degree of target magnetic force pump failure, obtain multiple correlation degree information;Multiple operating parameters in operating parameter set are input into the operating detection unit in magnetic force pump detection model, obtain first quality detection result;Collect the image information of the isolation component in the target magnetic force pump;Image information is input into the isolation detection unit in magnetic force pump detection model, obtain second quality detection result;Integrate first quality detection result and second quality detection result, obtain the quality detection result of target magnetic force pump.The technical problem that quality detection accuracy is insufficient, comprehensiveness is low for particle-resistant type magnetic force pump in prior art, in turn, cause the quality detection effect of particle-resistant type magnetic force pump is not good.

Description

Technical Field

[0001] This invention relates to the field of magnetic pump testing, specifically to an intelligent testing method and system for particle-resistant magnetic pumps. Background Technology

[0002] With the widespread application of particle-resistant magnetic drive pumps, their quality inspection has attracted considerable attention. Current technologies suffer from insufficient accuracy and comprehensiveness in quality inspection of particle-resistant magnetic drive pumps, resulting in poor quality inspection outcomes. Therefore, researching and designing an optimized quality inspection method for particle-resistant magnetic drive pumps is of significant practical importance. Summary of the Invention

[0003] This application provides an intelligent detection method and system for particle-resistant magnetic pumps. It solves the technical problems of insufficient accuracy and comprehensiveness in the quality detection of particle-resistant magnetic pumps in existing technologies, resulting in poor quality detection effects. It achieves the technical effect of improving the accuracy and comprehensiveness of quality detection for particle-resistant magnetic pumps through multi-dimensional quality analysis, thereby enhancing the overall quality detection effect.

[0004] In view of the above problems, this application provides a method and system for intelligent detection of particle-resistant magnetic pumps.

[0005] In a first aspect, this application provides an intelligent detection method for particle-resistant magnetic pumps, wherein the method is applied to an intelligent detection system for particle-resistant magnetic pumps. The method includes: based on the Internet of Things, acquiring a set of operating parameters including multiple categories of operating parameters through a detection device pre-set inside the target magnetic pump to be detected, wherein the detection device includes multiple sensors, and the multiple categories of operating parameters include pressure, vibration signal, temperature, and flow rate; analyzing the correlation between changes in the multiple categories of operating parameters and the failure of the target magnetic pump at a preset time point before the target magnetic pump fails within a preset historical time range, and obtaining multiple correlation information; inputting the multiple operating parameters in the set of operating parameters into an operating detection unit within a magnetic pump detection model to obtain a first quality detection result, wherein the operating detection unit is constructed based on the multiple correlation information; acquiring image information of an isolation component inside the target magnetic pump, wherein the isolation component is a component that isolates impurity particles; inputting the image information into an isolation detection unit within the magnetic pump detection model to obtain a second quality detection result; and integrating the first quality detection result and the second quality detection result to obtain a quality detection result of the target magnetic pump.

[0006] Secondly, this application also provides a particle-resistant magnetic pump intelligent detection system, wherein the system includes: a parameter acquisition module, which is used to acquire a set of operating parameters including multiple operating parameter categories based on the Internet of Things and through a detection device pre-set inside the target magnetic pump to be detected, wherein the detection device includes multiple sensors and the multiple operating parameter categories include pressure, vibration signal, temperature, and flow rate; a fault correlation analysis module, which is used to analyze the correlation between changes in the multiple operating parameter categories and the failure of the target magnetic pump based on historical operating parameters of the target magnetic pump at a preset time point before the failure within a preset historical time range, and obtain multiple correlation information; a first quality detection module, the first... The quality detection module is used to input multiple operating parameters from the set of operating parameters into the operating detection unit within the magnetic pump detection model to obtain a first quality detection result, wherein the operating detection unit is constructed based on the multiple correlation information; the image acquisition module is used to acquire image information of the isolation component within the target magnetic pump, wherein the isolation component is a component that isolates impurity particles; the second quality detection module is used to input the image information into the isolation detection unit within the magnetic pump detection model to obtain a second quality detection result; the quality detection result acquisition module is used to integrate the first quality detection result and the second quality detection result to obtain the quality detection result of the target magnetic pump.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The detection device collects data on various operating parameters of the target magnetic pump to obtain a set of operating parameters. Based on the historical operating parameters of the target magnetic pump at a preset time point before the failure within a preset historical time range, the correlation between changes in various operating parameter categories and the failure of the target magnetic pump is analyzed to obtain multiple correlation information. The set of operating parameters is input into the operation detection unit within the magnetic pump detection model to obtain a first quality detection result. The image information of the isolation components inside the target magnetic pump is input into the isolation detection unit within the magnetic pump detection model to obtain a second quality detection result. Combining the second quality detection result, the final quality detection result of the target magnetic pump is obtained. This achieves the technical effect of improving the accuracy and comprehensiveness of the quality detection of particle-resistant magnetic pumps through multi-dimensional quality analysis, thereby enhancing the quality detection effect of particle-resistant magnetic pumps.

[0009] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings of the embodiments of this disclosure will be briefly described below. Obviously, the drawings described below only relate to some embodiments of this disclosure and are not intended to limit this disclosure.

[0011] Figure 1 This is a flowchart illustrating the intelligent detection method for particle-resistant magnetic pumps according to this application.

[0012] Figure 2 This is a flowchart illustrating the process of obtaining the set of operating parameters in the intelligent detection method for particle-resistant magnetic pumps according to this application.

[0013] Figure 3 This is a schematic diagram of the structure of a particle-resistant magnetic pump intelligent detection system according to this application.

[0014] Figure labeling: Parameter acquisition module 11, Fault correlation analysis module 12, First quality inspection module 13, Image acquisition module 14, Second quality inspection module 15, Quality inspection result acquisition module 16. Detailed Implementation

[0015] This application provides an intelligent detection method and system for particle-resistant magnetic pumps. It solves the technical problems of insufficient accuracy and comprehensiveness in the quality detection of particle-resistant magnetic pumps in existing technologies, resulting in poor quality detection effects. By performing multi-dimensional quality analysis on particle-resistant magnetic pumps, it achieves the technical effect of improving the accuracy and comprehensiveness of quality detection, and ultimately enhancing the quality detection results.

[0016] Example 1

[0017] Please see the appendix Figure 1 This application provides a method for intelligent detection of particle-resistant magnetic pumps, wherein the method is applied to an intelligent detection system for particle-resistant magnetic pumps, and the method specifically includes the following steps:

[0018] Step S100: Based on the Internet of Things, a set of operating parameters including multiple categories is collected through a detection device pre-set inside the target magnetic pump to be tested. The detection device includes multiple sensors, and the multiple categories of operating parameters include pressure, vibration signal, temperature, and flow rate.

[0019] Further details are attached. Figure 2 As shown, step S100 of this application further includes:

[0020] Step S110: Based on the Internet of Things, the pressure, vibration signal, temperature and flow rate of the target magnetic pump under the current operating conditions are collected by a detection device pre-set inside the target magnetic pump to be tested.

[0021] Step S120: Process the vibration signal for vibration frequency, vibration peak value and vibration variance, and combine it with the pressure, temperature and flow rate to obtain the set of operating parameters.

[0022] Specifically, based on the Internet of Things (IoT), a pre-installed detection device within the target magnetic pump collects data on the pump's pressure, vibration signals, temperature, and flow rate under current operating conditions. The vibration signals are processed for vibration frequency, peak value, and variance. Combined with the pressure, temperature, and flow rate, a set of operating parameters is obtained. The target magnetic pump is any particle-resistant magnetic pump that can be intelligently tested using this particle-resistant magnetic pump intelligent detection system. The detection device includes multiple sensors, including pressure sensors, vibration signal sensors, temperature sensors, and flow sensors. The set of operating parameters includes the target magnetic pump's pressure, vibration signals, temperature, and flow rate under current operating conditions. The vibration signals include vibration frequency, peak value of the vibration signal amplitude, and vibration variance. The vibration variance includes the variance corresponding to the vibration amplitude. This achieves the technical effect of determining the target magnetic pump's set of operating parameters under current conditions, laying the foundation for subsequent quality testing of the target magnetic pump.

[0023] Step S200: Based on the historical operating parameters of the target magnetic pump at a preset time point before the failure within a preset historical time range, analyze the correlation between the changes in the various operating parameter categories and the failure of the target magnetic pump, and obtain multiple correlation information.

[0024] Furthermore, step S200 of this application also includes:

[0025] Step S210: Obtain the standard range of multiple operating parameters for the target magnetic pump under the current operating conditions for the various operating parameter categories;

[0026] Step S220: Obtain historical operating parameters of the target magnetic pump at preset time points before multiple failures within a preset historical time range, and obtain multiple sets of historical operating parameters;

[0027] Specifically, based on various operating parameter categories, multiple standard ranges of operating parameters for the target magnetic pump are obtained. Based on preset time points prior to multiple failures within a preset historical time range, historical operating parameters of the target magnetic pump are collected, resulting in multiple sets of historical operating parameters. Each standard range of operating parameters includes pre-defined standard parameter range information for each type of operating parameter under normal operating conditions of the target magnetic pump. This parameter range information includes the interval information consisting of the minimum and maximum standard values ​​of the operating parameters for each type of operating parameter.

[0028] The preset historical time range includes preset time points prior to multiple failures. Each preset time point prior to a failure includes a preset historical time point prior to each failure of the target magnetic pump. For example, the preset historical time range can be the past month or year, and the preset time point can be 5 seconds prior to each failure. Each set of historical operating parameters includes multiple historical operating parameters corresponding to each operating parameter category under the preset time points prior to multiple failures within the preset historical time range.

[0029] Step S230: Based on the multiple standard ranges of operating parameters and the multiple sets of historical operating parameters, analyze the correlation between changes in the operating parameters of the multiple operating parameter categories and the failure of the target magnetic pump, and obtain the multiple correlation information.

[0030] Furthermore, step S230 of this application also includes:

[0031] Step S231: Calculate the proportion of historical operating parameters in the multiple historical operating parameter sets that exceed the standard range of the multiple operating parameters, and obtain multiple historical parameter abnormality proportion parameter sets;

[0032] Step S232: Calculate the mean of the multiple historical parameter aberration ratio parameter sets respectively to obtain multiple average parameter aberration ratio parameters;

[0033] Step S233: Count the number of times each of the various operating parameter categories exceeds the corresponding operating parameter standard range within the multiple historical operating parameter sets, and obtain multiple abnormality counts;

[0034] Step S234: Calculate the reciprocal of the multiple average parameter abnormality ratio parameters, and calculate the product with the multiple abnormality counts to obtain the multiple correlation information.

[0035] Specifically, based on multiple standard ranges for operating parameters, the proportion of historical operating parameters exceeding their corresponding standard ranges within multiple sets of historical operating parameters is calculated to obtain multiple sets of historical parameter anomaly ratio parameters. These sets of historical parameter anomaly ratio parameters correspond to the multiple sets of historical operating parameters. Each set of historical parameter anomaly ratio parameters includes multiple historical parameter anomaly ratio parameters corresponding to each set of historical operating parameters. For example, when obtaining multiple sets of historical parameter anomaly ratio parameters, each historical operating parameter in each of the multiple sets of historical operating parameters is input into the anomaly ratio evaluation formula to obtain the historical parameter anomaly ratio parameter corresponding to each historical operating parameter in the multiple sets of historical operating parameters.

[0036] For example, the abnormality ratio assessment formula includes

[0037] Where f(x) is the output historical parameter deviation ratio parameter, and X is the input historical operating parameter. a X is the minimum standard value of the operating parameter within the standard range corresponding to the input historical operating parameters. b This refers to the maximum standard value of the operating parameter within the standard range corresponding to the historical operating parameters input.

[0038] Furthermore, the mean of multiple historical parameter anomaly ratio parameter sets is calculated, resulting in multiple average parameter anomaly ratio parameters. Each average parameter anomaly ratio parameter includes the average value of multiple historical parameter anomaly ratio parameters within each historical parameter anomaly ratio parameter set. Next, multiple historical operating parameter sets are statistically analyzed to obtain multiple anomaly counts. Each anomaly count includes the number of times each operating parameter category exceeds the corresponding operating parameter standard range. That is, each anomaly count includes the number of non-zero historical parameter anomaly ratio parameters within each historical parameter anomaly ratio parameter set. Then, the reciprocals of the multiple average parameter anomaly ratio parameters are multiplied by the corresponding multiple anomaly counts to obtain multiple correlation information. Each correlation information includes the product of the reciprocal of each average parameter anomaly ratio parameter and the corresponding anomaly count. The smaller the average parameter anomaly ratio parameter and the more anomaly counts, the more prone the corresponding operating parameter category is to anomaly. Furthermore, even a small deviation in the operating parameters of this category can lead to a failure of the target magnetic pump, thus the stronger the correlation between this operating parameter category and the target magnetic pump's failure. This achieves the technical effect of improving the reliability of quality testing of the target magnetic pump by analyzing the failure correlation of multiple operating parameter categories through multiple standard ranges and historical operating parameter sets, obtaining accurate correlation information across multiple categories.

[0039] Step S300: Input multiple operating parameters from the set of operating parameters into the operating detection unit within the magnetic pump detection model to obtain a first quality detection result, wherein the operating detection unit is constructed based on the multiple correlation information;

[0040] Furthermore, step S300 of this application also includes:

[0041] Step S310: Obtain a set of multiple sample operating parameters for the various operating parameter categories;

[0042] Step S320: Obtain the failure probability of the target magnetic pump when the various operating parameter categories are in different specific operating parameters, and obtain multiple sample failure probability sets;

[0043] Specifically, samples of the target magnetic pump are collected based on multiple operating parameter categories to obtain multiple sets of sample operating parameters and multiple sets of sample failure probabilities. The multiple sets of sample operating parameters correspond to multiple operating parameter categories. Each set of sample operating parameters includes multiple sample operating parameters. These multiple sample operating parameters include historical operating parameter information corresponding to each operating parameter category. Each set of sample failure probabilities includes multiple failure probabilities corresponding to the multiple sample operating parameters within each set. For example, when obtaining multiple failure probabilities, multiple monitoring data are collected from the target magnetic pump based on a preset historical time period to obtain multiple monitoring data. Each monitoring data includes data on whether the target magnetic pump failed under different specific operating parameters within multiple operating parameter categories. Based on the multiple monitoring data, the ratio between the number of times the target magnetic pump failed under each specific sample operating parameter and the total number of monitoring times for that specific sample operating parameter is calculated to obtain multiple failure probabilities. Different sample operating parameters within each operating parameter category correspond to different failure probabilities.

[0044] Step S330: Using the first sample operating parameter set and the first sample fault probability set within the multiple sample operating parameter sets and multiple sample fault probability sets as construction data, construct the first operating detection branch corresponding to the first operating parameter category;

[0045] Furthermore, step S330 of this application also includes:

[0046] Step S331: Construct the network structure of the first running detection branch based on the BP neural network;

[0047] Step S332: Identify the first sample running parameter set and the first sample failure probability set as the first constructed dataset;

[0048] Step S333: Use the first constructed dataset to perform supervised training, verification and testing on the first running detection branch to obtain the first running detection branch that meets the preset conditions.

[0049] Specifically, a first operating parameter category is obtained by randomly selecting from multiple operating parameter categories. Then, multiple sample operating parameter sets and multiple sample failure probability sets are matched according to this first operating parameter category to obtain a first sample operating parameter set and a first sample failure probability set. The first operating parameter category can be any one of the multiple operating parameter categories. The first sample operating parameter set and the first sample failure probability set include the sample operating parameter set and sample failure probability set corresponding to the first operating parameter category from among the multiple sample operating parameter sets and multiple sample failure probability sets.

[0050] A BP neural network is a multi-layer feedforward neural network trained using the backpropagation algorithm. A BP neural network consists of an input layer, multiple layers of neurons, and an output layer. It can perform both forward and backward computation. During forward computation, the input information is processed layer by layer from the input layer through multiple neurons before reaching the output layer; the state of each neuron only affects the state of the next layer. If the desired output is not obtained at the output layer, backward computation is initiated, returning the error signal along the original connection path. By modifying the weights of each neuron, the error signal is minimized. The BP neural network is used as the network structure for the first running detection branch. That is, the network structure for the first running detection branch includes an input layer, hidden layers, and an output layer.

[0051] Further, the first sample operating parameter set and the first sample failure probability set are labeled to obtain a first constructed dataset. 70% of the data in the first constructed dataset is randomly divided into a training set, 20% into a test set, and 10% into a validation set. Supervised training is performed on the training set using a BP neural network to obtain a first operating detection branch. The test set is used as input to the first operating detection branch to update its parameters. The validation set is used as input to the first operating detection branch to validate it, obtaining a first operating detection branch that meets preset conditions. The first constructed dataset includes the first sample operating parameter set and the first sample failure probability set. The preset conditions include a pre-defined output accuracy threshold corresponding to the first operating detection branch. The first operating detection branch includes an input layer, a hidden layer, and an output layer. The first operating detection branch corresponds to a first operating parameter category. The first operating detection branch has the function of intelligently analyzing and matching the failure probabilities of the input operating parameters of the first operating parameter category. The technology achieves the effect of constructing a first operation detection branch that meets preset conditions through a BP neural network, thereby improving the accuracy of the constructed operation detection unit.

[0052] Step S340: Continue to construct multiple operation detection branches within the operation detection unit, and construct a weighted calculation branch based on the multiple correlation information;

[0053] Step S350: Input multiple operating parameters from the set of operating parameters into the multiple operating detection branches to obtain multiple fault probability information, input them into the weighted calculation branch, and calculate the multiple fault probability information according to the magnitude of the multiple correlation information to obtain the total fault probability, which is used as the first quality detection result.

[0054] Specifically, based on multiple operating parameter categories, multiple operating parameter sets and multiple sample fault probability sets are constructed to form multiple operating detection branches. A weighted calculation branch is constructed based on multiple correlation information. The multiple operating detection branches and the weighted calculation branch are connected to obtain an operating detection unit. The magnetic pump detection model includes an operating detection unit. The operating detection unit includes multiple operating detection branches and a weighted calculation branch. The multiple operating detection branches have a one-to-one correspondence with multiple operating parameter categories. The multiple operating detection branches include a first operating detection branch. The construction method of the multiple operating detection branches and the first operating detection branch is the same, and will not be repeated here for the sake of brevity. The weighted calculation branch includes multiple fault weight values ​​corresponding to multiple operating parameter categories. Preferably, multiple correlation information is summed to obtain total correlation information. The ratios of each correlation information to the total correlation information are calculated to obtain multiple fault weight values. Each fault weight value includes the ratio between each correlation information and the total correlation information.

[0055] Furthermore, according to various operating parameter categories, multiple operating parameters within the operating parameter set are input into the corresponding operating detection branch to obtain multiple fault probability information. These multiple fault probability information are then input into a weighted calculation branch, where multiple fault weight values ​​within the weighted calculation branch are used to perform weighted calculations to obtain the total fault probability, which is then output as the first quality inspection result. Each operating detection branch has the function of intelligently analyzing and matching the fault probabilities of the input operating parameter categories. For example, when obtaining the total fault probability, multiple fault probability information is multiplied by corresponding multiple fault weight values ​​to obtain multiple weighted fault probability information. The sum of these weighted fault probability information is output as the total fault probability. The first quality inspection result includes the total fault probability. This achieves the technical effect of improving the accuracy of quality inspection of particle-resistant magnetic pumps by intelligently analyzing the operating parameter set through the operating detection unit to obtain accurate first quality inspection results.

[0056] Step S400: Acquire image information of the isolation component inside the target magnetic pump, wherein the isolation component is a component that isolates impurity particles;

[0057] Step S500: Input the image information into the isolation detection unit within the magnetic pump detection model to obtain the second quality detection result;

[0058] Furthermore, step S500 of this application also includes:

[0059] Step S510: Obtain a set of sample image information of the isolation component, and perform wear analysis on the isolation component to obtain a set of sample wear analysis results;

[0060] Step S520: Using the sample image information set and the sample wear analysis result set as construction data, the isolation detection unit is constructed and trained based on the convolutional neural network, and combined with the running detection unit, the magnetic pump detection model is obtained;

[0061] Step S530: Input the image information into the isolation detection unit to obtain the wear analysis result, which is used as the second quality detection result.

[0062] Step S600: Integrate the first quality inspection result and the second quality inspection result to obtain the quality inspection result of the target magnetic pump.

[0063] Specifically, image acquisition is performed on the isolation components of the target magnetic pump to obtain image information. This image information includes image data corresponding to the isolation components of the target magnetic pump. The isolation components are those that isolate impurity particles from the target magnetic pump. For example, the isolation components include the isolation sleeve and the pump body of the target magnetic pump.

[0064] Furthermore, based on the isolation components, sample image information sets are collected, and the wear degree of the isolation components is identified in the sample image information sets to obtain a sample wear analysis result set. Then, the sample image information set and the sample wear analysis result set are used as construction data. Based on a convolutional neural network, the construction data is continuously self-trained until convergence to obtain an isolation detection unit. Subsequently, the image information is input into the isolation detection unit to obtain wear analysis results, which are used as the second quality detection result. Combined with the first quality detection result, the quality detection result of the target magnetic pump is obtained. The sample image information set includes multiple sample image information. The multiple sample image information includes multiple historical image information of multiple sample isolation components. The multiple sample isolation components include the isolation component of the target magnetic pump, and multiple isolation components of the same type. The sample wear analysis result set includes multiple wear coefficients of multiple sample isolation components corresponding to the multiple sample image information. The larger the wear coefficient of the sample isolation component, the more severe the wear degree of the corresponding sample isolation component. The convolutional neural network is a type of feedforward neural network that includes convolution calculations and has a deep structure. Convolutional neural networks possess representation learning capabilities, enabling translation-invariant classification of input information according to their hierarchical structure. The isolation detection unit comprises an input layer, a hidden layer, and an output layer. This unit has the function of intelligently analyzing the corresponding input image information and matching the wear coefficient of the isolation components. The magnetic pump detection model includes a running detection unit and an isolation detection unit. The wear analysis result includes the wear coefficient of the isolation components corresponding to the image information. The second quality detection result includes the wear analysis result. The quality detection result of the target magnetic pump includes the first quality detection result and the second quality detection result. This achieves the technical effect of improving the quality detection effect of particle-resistant magnetic pumps by performing multi-dimensional quality analysis of the target magnetic pump through the magnetic pump detection model.

[0065] In summary, the intelligent detection method for particle-resistant magnetic pumps provided in this application has the following technical advantages:

[0066] 1. Data collection of various operating parameters of the target magnetic pump is performed using a detection device to obtain a set of operating parameters. Based on the historical operating parameters of the target magnetic pump at a preset time point before the pump's failure within a preset historical time range, the correlation between changes in various operating parameter categories and the pump's failure is analyzed to obtain multiple correlation information. The set of operating parameters is input into the operating detection unit within the magnetic pump detection model to obtain a first quality detection result. Image information of the isolation components within the target magnetic pump is input into the isolation detection unit within the same model to obtain a second quality detection result. Combining the second quality detection result, the final quality detection result of the target magnetic pump is obtained. This achieves the technical effect of improving the accuracy and comprehensiveness of particle-resistant magnetic pump quality detection and enhancing its quality detection performance through multi-dimensional quality analysis.

[0067] 2. By using multiple standard ranges of operating parameters and multiple sets of historical operating parameters, fault correlation analysis is performed on various categories of operating parameters to obtain accurate correlation information, thereby improving the reliability of quality testing of the target magnetic pump.

[0068] 3. By performing intelligent fault probability analysis on the set of operating parameters through the operation detection unit, accurate first quality detection results are obtained, thereby improving the accuracy of quality detection of particle-resistant magnetic pumps.

[0069] Example 2

[0070] Based on the same inventive concept as the intelligent detection method for particle-resistant magnetic pumps described in the foregoing embodiments, this invention also provides an intelligent detection system for particle-resistant magnetic pumps. Please refer to the appendix. Figure 3 The system includes:

[0071] The parameter acquisition module 11 is used to acquire a set of operating parameters, including multiple categories of operating parameters, through a detection device pre-set inside the target magnetic pump to be tested, based on the Internet of Things. The detection device includes multiple sensors, and the multiple categories of operating parameters include pressure, vibration signal, temperature, and flow rate.

[0072] The fault correlation analysis module 12 is used to analyze the correlation between changes in various operating parameter categories and the failure of the target magnetic pump based on the historical operating parameters of the target magnetic pump at a preset time point before the failure within a preset historical time range, and to obtain multiple correlation information.

[0073] The first quality detection module 13 is used to input multiple operating parameters from the set of operating parameters into the operating detection unit in the magnetic pump detection model to obtain a first quality detection result. The operating detection unit is constructed based on the multiple correlation information.

[0074] Image acquisition module 14, the image acquisition module 14 is used to acquire image information of the isolation component inside the target magnetic pump, wherein the isolation component is a component that isolates impurity particles;

[0075] The second quality detection module 15 is used to input the image information into the isolation detection unit in the magnetic pump detection model to obtain the second quality detection result.

[0076] The quality inspection result acquisition module 16 is used to integrate the first quality inspection result and the second quality inspection result to obtain the quality inspection result of the target magnetic pump.

[0077] Furthermore, the system also includes:

[0078] The first execution module is used to collect the pressure, vibration signal, temperature and flow rate of the target magnetic pump under the current operating conditions through a detection device pre-set inside the target magnetic pump to be detected, based on the Internet of Things.

[0079] The operating parameter set acquisition module is used to process the vibration signal by vibration frequency, vibration peak value and vibration variance, and combine it with the pressure, temperature and flow rate to obtain the operating parameter set.

[0080] Furthermore, the system also includes:

[0081] A standard range acquisition module is used to acquire the standard range of multiple operating parameters of the target magnetic pump under the current operating conditions for the various operating parameter categories.

[0082] The historical operating parameter acquisition module is used to acquire the historical operating parameters of the target magnetic pump at a preset time point before multiple failures within a preset historical time range, and to obtain multiple sets of historical operating parameters.

[0083] The correlation information determination module is used to analyze the correlation between changes in operating parameters of the multiple operating parameter categories and the failure of the target magnetic pump based on the multiple standard ranges of operating parameters and the multiple historical operating parameter sets, and to obtain the multiple correlation information.

[0084] Furthermore, the system also includes:

[0085] The second execution module is used to calculate the proportion of historical operating parameters in the multiple historical operating parameter sets that exceed the standard range of the multiple operating parameters, and obtain multiple historical parameter abnormality proportion parameter sets.

[0086] The average parameter aberration ratio parameter calculation module is used to calculate the mean of the multiple historical parameter aberration ratio parameter sets respectively, and obtain multiple average parameter aberration ratio parameters.

[0087] An error count module is used to count the number of times each of the various operating parameter categories exceeds the corresponding operating parameter standard range within the multiple historical operating parameter sets, thereby obtaining multiple error counts.

[0088] The third execution module is used to calculate the reciprocal of the multiple average parameter abnormality ratio parameters, and calculate the product with the multiple abnormality counts to obtain the multiple correlation information.

[0089] Furthermore, the system also includes:

[0090] A sample operation parameter acquisition module, which is used to acquire multiple sets of sample operation parameters of the various operation parameter categories;

[0091] The fault probability acquisition module is used to acquire the fault probability of the target magnetic pump when the multiple operating parameter categories are in different specific operating parameters, and to acquire multiple sample fault probability sets.

[0092] The fourth execution module is used to construct a first operation detection branch corresponding to the first operation parameter category by using the first sample operation parameter set and the first sample failure probability set within the multiple sample operation parameter sets and multiple sample failure probability sets as construction data.

[0093] A weighted calculation branch acquisition module is used to further construct and obtain multiple operation detection branches within the operation detection unit, and construct weighted calculation branches based on the multiple correlation information.

[0094] The fifth execution module is used to input multiple operating parameters from the set of operating parameters into the multiple operating detection branches to obtain multiple fault probability information, input them into the weighted calculation branch, and calculate the multiple fault probability information according to the magnitude of the multiple correlation information to obtain the total fault probability, which is used as the first quality detection result.

[0095] Furthermore, the system also includes:

[0096] A network structure construction module is used to construct the network structure of the first running detection branch based on a BP neural network.

[0097] The data identification module is used to identify the first sample operating parameter set and the first sample failure probability set as the first constructed dataset;

[0098] The sixth execution module is used to perform supervised training, verification and testing on the first running detection branch using the first constructed dataset, so as to obtain the first running detection branch that meets the preset conditions.

[0099] Furthermore, the system also includes:

[0100] The sample wear analysis module is used to acquire a set of sample image information of the isolation component, analyze the wear condition of the isolation component, and obtain a set of sample wear analysis results.

[0101] The seventh execution module is used to construct and train the isolation detection unit based on the sample image information set and the sample wear analysis result set as construction data, and combine the operation detection unit to obtain the magnetic pump detection model.

[0102] The second quality inspection result output module is used to input the image information into the isolation detection unit to obtain wear analysis results, which are used as the second quality inspection result.

[0103] The particle-resistant magnetic pump intelligent detection system provided in this embodiment of the invention can execute the particle-resistant magnetic pump intelligent detection method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0104] The modules included are divided according to functional logic, but are not limited to the above division, as long as they can achieve the corresponding functions; in addition, the specific names of each functional module are only for easy distinction between each other and are not used to limit the scope of protection of this invention.

[0105] This application provides an intelligent detection method for particle-resistant magnetic pumps. The method is applied to an intelligent detection system for particle-resistant magnetic pumps. The method includes: collecting data on various operating parameters of a target magnetic pump using a detection device to obtain a set of operating parameters; analyzing the correlation between changes in various operating parameter categories and the failure of the target magnetic pump based on historical operating parameters at a preset time point before the pump's failure within a preset historical time range, obtaining multiple correlation information; inputting the set of operating parameters into an operating detection unit within a magnetic pump detection model to obtain a first quality detection result; inputting image information of the isolation components within the target magnetic pump into an isolation detection unit within the magnetic pump detection model to obtain a second quality detection result; and combining the second quality detection result to obtain the final quality detection result of the target magnetic pump. This method solves the technical problem of insufficient accuracy and comprehensiveness in the existing quality detection of particle-resistant magnetic pumps, resulting in poor quality detection performance. It achieves the technical effect of improving the accuracy and comprehensiveness of particle-resistant magnetic pump quality detection through multi-dimensional quality analysis, thereby enhancing the overall quality detection performance.

[0106] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. A method for intelligent detection of particle-resistant magnetic pumps, characterized in that, The method includes: Based on the Internet of Things, a set of operating parameters including various categories is collected through a detection device pre-set inside the target magnetic pump to be tested. The detection device includes various sensors, and the various categories of operating parameters include pressure, vibration signal, temperature, and flow rate. Based on the historical operating parameters of the target magnetic pump at a preset time point before the failure within a preset historical time range, analyze the correlation between changes in various operating parameter categories and the failure of the target magnetic pump to obtain multiple correlation information. Multiple operating parameters from the set of operating parameters are input into the operating detection unit within the magnetic pump detection model to obtain a first quality detection result, wherein the operating detection unit is constructed based on the multiple correlation information. Image information of the isolation component inside the target magnetic pump is acquired, wherein the isolation component is a component that isolates impurity particles; The image information is input into the isolation detection unit within the magnetic pump detection model to obtain a second quality detection result. By integrating the first quality inspection result and the second quality inspection result, the quality inspection result of the target magnetic pump is obtained; Based on the historical operating parameters of the target magnetic pump at a preset time point before its failure within a preset historical time range, analyze the correlation between changes in various operating parameter categories and the failure of the target magnetic pump, including: Obtain the standard range of multiple operating parameters for the target magnetic pump under the current operating conditions for various operating parameter categories; Obtain historical operating parameters of the target magnetic pump at preset time points before multiple failures within a preset historical time range, and obtain multiple sets of historical operating parameters. Based on the multiple standard ranges of operating parameters and the multiple sets of historical operating parameters, analyze the correlation between changes in operating parameters of the multiple categories of operating parameters and the failure of the target magnetic pump, and obtain the multiple correlation information. Each operating parameter standard range includes pre-defined standard parameter range information for the target magnetic pump under normal operation for each type of operating parameter. This parameter range information includes the interval information consisting of the minimum and maximum standard values ​​of the operating parameters for each type of operating parameter.

2. The method according to claim 1, characterized in that, Based on the Internet of Things (IoT), a set of operating parameters, including various categories, is collected through a pre-installed detection device inside the target magnetic pump to be tested. Based on the Internet of Things, the pressure, vibration signal, temperature and flow rate of the target magnetic pump under the current operating conditions are collected by a detection device pre-installed inside the target magnetic pump to be tested. The vibration signal is processed to obtain the vibration frequency, vibration peak value and vibration variance, and combined with the pressure, temperature and flow rate to obtain the set of operating parameters.

3. The method according to claim 1, characterized in that, Based on the multiple standard ranges of operating parameters and the multiple sets of historical operating parameters, analyze the correlation between changes in operating parameters of the multiple operating parameter categories and the failure of the target magnetic pump, and obtain the multiple correlation information, including: Calculate the proportion of historical operating parameters exceeding the standard range of the multiple historical operating parameter sets to obtain multiple sets of historical parameter abnormality proportion parameters. Calculate the mean of each of the multiple historical parameter aberration ratio parameter sets to obtain multiple average parameter aberration ratio parameters; The number of times each of the various operating parameter categories exceeded the corresponding standard range within the multiple historical operating parameter sets is counted to obtain multiple abnormality counts; Calculate the reciprocal of the multiple average parameter abnormality ratio parameters, and multiply it by the multiple abnormality frequencies to obtain the multiple correlation information.

4. The method according to claim 1, characterized in that, Multiple operating parameters from the set of operating parameters are input into the operating detection unit within the magnetic pump detection model to obtain a first quality detection result, including: Obtain multiple sample sets of operating parameters for the various operating parameter categories; The failure probability of the target magnetic pump is obtained when the various operating parameter categories are in different specific operating parameters, and multiple sample failure probability sets are obtained. Using the first sample operating parameter set and the first sample fault probability set within the multiple sample operating parameter sets and multiple sample fault probability sets as construction data, a first operating detection branch corresponding to the first operating parameter category is constructed. Continue to construct multiple operation detection branches within the operation detection unit, and construct a weighted calculation branch based on the multiple correlation information; Multiple operating parameters from the set of operating parameters are input into the multiple operating detection branches to obtain multiple fault probability information. These are then input into the weighted calculation branch, where the multiple fault probability information is weighted and calculated according to the magnitude of the multiple correlation information to obtain the total fault probability, which is used as the first quality detection result.

5. The method according to claim 4, characterized in that, Using the first sample operating parameter set and the first sample failure probability set from the multiple sample operating parameter sets and multiple sample failure probability sets as construction data, a first operating detection branch corresponding to the first operating parameter category is constructed, including: Based on the BP neural network, the network structure of the first running detection branch is constructed; The first sample operating parameter set and the first sample failure probability set are labeled with data to form the first constructed dataset; The first running detection branch is trained, validated, and tested using the first constructed dataset to obtain a first running detection branch that meets preset conditions.

6. The method according to claim 1, characterized in that, The image information is input into the isolation detection unit within the magnetic pump detection model to obtain a second quality detection result, including: A set of sample image information of the isolation component is obtained, and the wear condition of the isolation component is analyzed to obtain a set of sample wear analysis results; Using the sample image information set and the sample wear analysis result set as construction data, the isolation detection unit is constructed and trained based on the convolutional neural network, and combined with the running detection unit, the magnetic pump detection model is obtained; The image information is input into the isolation detection unit to obtain wear analysis results, which are used as the second quality detection results.

7. A particle-resistant magnetic pump intelligent detection system, characterized in that, The system is used to perform the method according to any one of claims 1 to 6, the system comprising: The parameter acquisition module is used to acquire a set of operating parameters, including multiple categories of operating parameters, through a detection device pre-set inside the target magnetic pump to be tested, based on the Internet of Things. The detection device includes multiple sensors, and the multiple categories of operating parameters include pressure, vibration signal, temperature, and flow rate. The fault correlation analysis module is used to analyze the correlation between changes in various operating parameter categories and the failure of the target magnetic pump based on the historical operating parameters of the target magnetic pump at a preset time point before the failure within a preset historical time range, and to obtain multiple correlation information. A standard range acquisition module is used to acquire the standard range of multiple operating parameters of the target magnetic pump under the current operating conditions for the various operating parameter categories. The historical operating parameter acquisition module is used to acquire the historical operating parameters of the target magnetic pump at a preset time point before multiple failures within a preset historical time range, and to obtain multiple sets of historical operating parameters. The correlation information determination module is used to analyze the correlation between changes in the operating parameters of the multiple operating parameter categories and the failure of the target magnetic pump based on the multiple standard ranges of operating parameters and the multiple historical operating parameter sets, and to obtain the multiple correlation information. The first quality detection module is used to input multiple operating parameters from the set of operating parameters into the operating detection unit within the magnetic pump detection model to obtain a first quality detection result. The operating detection unit is constructed based on the multiple correlation information. An image acquisition module is used to acquire image information of the isolation component inside the target magnetic pump, wherein the isolation component is a component that isolates impurity particles; The second quality detection module is used to input the image information into the isolation detection unit within the magnetic pump detection model to obtain the second quality detection result. A quality inspection result acquisition module is used to integrate the first quality inspection result and the second quality inspection result to obtain the quality inspection result of the target magnetic pump.

Images