Dredge pump performance analysis method based on numerical simulation and model test

Abstract

The invention relates to the technical field of dredge pump performance analysis, and discloses a numerical simulation and model test-based dredge pump performance analysis method, which comprises the following steps of: planning an analysis process, and determining numerical simulation (dredge pump internal flow field simulation) and model test contents (actually measured performance data under different working conditions); determining working condition types according to design parameters and working conditions; generating a performance analysis process set and determining an execution sequence; generating a simulation parameter setting list and a test scheme list of each working condition; and associating the data to generate an overall scheme. The method further relates to working condition analysis boundary determination, working condition type subdivision, dynamic adjustment of the process, test rule optimization, database construction and the like. According to the method, comprehensive and accurate analysis of the dredge pump performance is achieved through multi-dimensional data integration and a dynamic mechanism, the analysis efficiency and reliability are improved, and the method is suitable for evaluation and optimization of the dredge pump performance.

Description

Technical Field

[0001] The invention relates to the technical field of mud pump performance analysis, and in particular to a mud pump performance analysis method based on numerical simulation and model test. Background Art

[0002] As a key fluid conveying equipment in the fields of mining, metallurgy, dredging, etc., the performance of mud pumps directly affects the operating efficiency, energy consumption and reliability of the system. As industrial production develops towards large-scale and intelligent directions, higher requirements are placed on the accurate performance analysis of mud pumps under different working conditions. Traditional mud pump performance analysis methods often have the following limitations: On the one hand, the analysis mode that solely relies on numerical simulation or model test is difficult to fully reflect the real performance of the mud pump. Although the internal flow field of the mud pump can be visualized and analyzed only through numerical simulation, the simulation results are greatly affected by factors such as boundary condition setting and grid division accuracy, and cannot fully reproduce the complex physical phenomena in actual operation; relying solely on model tests faces the problems of high cost, long cycle, and limited coverage of working conditions, especially for the performance test of mud pumps under extreme working conditions or new structures, the test difficulty and risk are significantly increased.

[0003] On the other hand, the existing analysis methods are not scientific enough in terms of working condition classification and performance evaluation. Traditional working condition classification is mostly based on a single flow or head indicator, and fails to comprehensively consider multi-dimensional data such as the efficiency, energy consumption, and operating environment of the mud pump, resulting in the analysis results being unable to accurately reflect the performance of the mud pump under actual complex working conditions. At the same time, the lack of a dynamic adjustment mechanism makes it difficult to adaptively optimize the analysis process and working condition type based on real-time operating data, which limits the timeliness and practicality of the analysis results.

[0004] In addition, in terms of data management and long-term performance tracking, traditional methods lack a systematic database construction and maintenance mechanism. A large number of simulation parameters, test data and performance analysis results have not been effectively integrated and utilized, making it difficult to achieve long-term tracking and comparative analysis of mud pump performance, which is not conducive to exploring the potential laws and optimization directions of mud pump performance changes.

[0005] With the rapid development of computer technology and sensor technology, the accuracy of numerical simulation algorithms has been continuously improved, and the automation degree and data acquisition capabilities of model tests have also been increasingly enhanced, providing technical support for the innovation of mud pump performance analysis methods. How to organically combine numerical simulation with model tests to establish a scientific, efficient, and dynamic mud pump performance analysis method to achieve a comprehensive and accurate evaluation of mud pump performance has become a technical problem that needs to be solved in this field. Summary of the invention

[0006] The purpose of the present invention is to provide a mud pump performance analysis method based on numerical simulation and model test to solve the problems raised in the above background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions: A method for analyzing the performance of a slurry pump based on numerical simulation and model tests, the method comprising: Planning the performance analysis process of the slurry pump based on the design parameters and working conditions of the slurry pump, determining the numerical simulation content and model test content required for the analysis, the numerical simulation content including the simulation of the internal flow field of the slurry pump, and the model test content including obtaining the measured performance data of the slurry pump under different working conditions; Determining different working condition types of the slurry pump according to the design parameters and working conditions of the slurry pump; Generating a set of slurry pump performance analysis processes according to the different working condition types, numerical simulation content and model test content of the slurry pump, and determining the execution order of the performance analysis process set based on the principle of optimal matching; Generating a list of simulation parameter settings for each working condition according to a preset analysis period, the different working condition types of the slurry pump and the requirements of the slurry pump performance index, and generating a list of test schemes for each working condition according to a preset test rule, the different working condition types of the slurry pump and the requirements of the slurry pump performance index; Associating the list of simulation parameter settings, the list of test schemes and the slurry pump performance analysis content for each working condition based on the execution order of the performance analysis process set to generate the overall scheme for the slurry pump performance analysis.

[0008] Preferably, the method plans the performance analysis process of the slurry pump based on the design parameters and working conditions of the slurry pump, and determines the numerical simulation and model test content required for the analysis, including: For the current working condition, determining the performance analysis boundary of the current working condition according to the historical performance data of the current working condition, the historical performance data of adjacent working conditions and a preset working condition division threshold; Extracting a subset of performance analysis data of the current working condition according to the performance analysis boundary and a preset time window; Determining the content of the numerical simulation and model test related data within the subset of performance analysis data as the slurry pump performance analysis content of the current working condition.

[0009] Preferably, the method determines different working condition types of the slurry pump according to the design parameters and working conditions of the slurry pump, including: For the current working condition, when the flow rate of the current working condition is greater than a preset flow rate threshold, determining the current working condition as a large flow rate working condition; When the flow rate of the current working condition is not greater than the preset flow rate threshold, classifying the head and efficiency data of the current working condition based on a first preset classification algorithm to obtain data of each performance category, and determining the central characteristics of the data of each performance category; Determine the performance potential level of the current working condition according to the working conditions of the current working condition; When the matching degree between the central features of the data of each performance category and the preset performance category is greater than the preset matching threshold, determine that the current working condition is a specific performance working condition; When the matching degree is not greater than the preset matching threshold, determine that the current working condition is an ordinary performance working condition.

[0010] Preferably, the content of the slurry pump performance analysis further includes the slurry pump structure data and the operating environment data. The different working condition types of the slurry pump include the high-efficiency working condition, the low-efficiency working condition, and the critical working condition. According to the different working condition types, numerical simulation content, and model test content of the slurry pump, a slurry pump performance analysis process set is generated, including: For the current working condition, classify the operating environment data of the current working condition based on the second preset classification algorithm to obtain the data of each operating environment category; for the current operating environment category data, screen the current operating environment category data based on the influence degree and change frequency of each environmental factor in the current operating environment category data to obtain the key environment subset of the current operating environment category data, and determine the influence label of the current operating environment category data according to the influence characteristics of the key environment subset and the mean value of the slurry pump performance index; According to whether the working condition type of the current working condition is a high-efficiency working condition; Preferably, after determining whether the working condition type of the current working condition is a high-efficiency working condition, the method further includes: If the working condition type of the current working condition is a high-efficiency working condition, match the influence label of the data of each operating environment category with the slurry pump performance index to generate a targeted optimization strategy; If the working condition type of the current working condition is a low-efficiency working condition, generate an improvement optimization strategy based on the preset improvement rules and performance improvement goals; If the working condition type of the current working condition is a critical working condition, generate a performance stability strategy according to the historical performance fluctuation and performance stability rules.

[0011] Preferably, after generating the slurry pump performance analysis process set, it further includes: Update the slurry pump performance analysis content of each working condition based on the real-time operation data of the slurry pump; Recalculate the working condition types of each working condition according to the updated slurry pump performance analysis content; Dynamically adjust the execution order of the performance analysis process set according to the recalculated working condition types.

[0012] Preferably, the preset test rules include: Dynamically adjust the sample size and the number of repetitions of the model test based on the confidence interval of the slurry pump historical test data; When the deviation between the measured performance data and the numerical simulation results exceeds the preset error threshold, the supplementary test module is automatically triggered.

[0013] Preferably, the preset test rules further include: Establish a calibration compensation mechanism for test equipment, and perform error compensation calibration on the test equipment based on the reference values of the standard flowmeter and pressure gauge before each test.

[0014] Preferably, after generating the overall scheme for the slurry pump performance analysis, the method further includes: Establish a slurry pump performance analysis database, and store data such as the simulation parameter setting list, test scheme list, slurry pump performance analysis content, and various working condition types in the overall scheme into this database; Regularly clean and maintain the data in the database, removing invalid data and duplicate data; Based on this database, conduct long-term tracking and comparative analysis on the slurry pump performance analysis results to evaluate the changing trend of the slurry pump performance over time.

[0015] Preferably, the second preset classification algorithm is a hierarchical clustering algorithm. The specific steps for classifying the operating environment data of the current working condition include: Calculate the similarity between each sample in the operating environment data; Gradually merge the samples into different categories according to the similarity to form a clustering tree; According to the preset number of clustering layers or clustering distance threshold, intercept appropriate categories from the clustering tree as the operating environment category data.

[0016] Compared with the prior art, the beneficial effects of the present invention are: At the level of analysis process planning, by combining the slurry pump design parameters and working conditions, systematically plan the numerical simulation and model test content, and clarify the analysis key points under different working condition types. For example, for the current working condition, determine the performance analysis boundary based on historical performance data, adjacent working condition data, and preset thresholds, extract the targeted data subset, ensure that the analysis content focuses on key issues, avoid redundant calculations and invalid tests, and effectively improve the analysis efficiency. At the same time, generate a set of performance analysis processes and determine the execution order based on the optimal matching principle, realizing the scientific scheduling of simulation and test processes, and reducing process conflicts and resource waste.

[0017] In terms of working condition classification and performance evaluation, the innovative working condition classification method breaks through the limitations of traditional single-index division. It initially distinguishes large-flow working conditions through a flow threshold, then uses the first preset classification algorithm to classify head and efficiency data, and combines performance potential level and matching degree judgment to subdivide working conditions into specific types such as high-efficiency, low-efficiency, and critical. This multi-dimensional classification method can more accurately reveal the performance characteristics of the slurry pump under different operating states. For example, for high-efficiency working conditions, an optimization strategy is generated by matching operating environment impact labels with performance indicators to further explore the potential of the equipment; for low-efficiency working conditions, an improvement strategy is formulated based on preset rules and improvement goals to target and solve performance shortboards; for critical working conditions, a stability strategy is generated based on historical fluctuations and stability rules to effectively prevent the risk of performance mutation.

[0018] In terms of data dynamic management and process optimization, a dynamic adjustment mechanism driven by real-time data is established. By updating analysis content, recalculating working condition types, and adjusting the process execution order with real-time operation data, the analysis method can adapt to changes in the operating state of the slurry pump. For example, when the deviation between measured data and simulation results exceeds the threshold, the supplementary test module is automatically triggered to ensure data reliability; the sample size and number of repetitions are dynamically adjusted based on the confidence interval of historical test data, reducing costs while ensuring test accuracy. In addition, the test equipment calibration compensation mechanism calibrates errors through the reference value of standard instruments, improving the accuracy of test data.

[0019] In terms of data storage and long-term tracking, a complete performance analysis database is constructed to realize the systematic storage and management of data such as simulation parameters, test schemes, analysis content, and working condition types. By regularly cleaning invalid data and duplicate data, the efficiency and availability of the database are guaranteed. Based on the long-term tracking and comparative analysis of the database, the change trend of the slurry pump performance over time can be clearly revealed, providing a scientific basis for equipment maintenance, upgrading, and transformation. For example, by comparing performance data from different periods, signs of equipment performance degradation can be detected in advance, and a preventive maintenance plan can be formulated to reduce the risk of downtime.

[0020] In terms of technology integration and application expansion, the advantages of flow field analysis in numerical simulation are combined with the reliability of measured data in model tests to form a complementary analysis system. Numerical simulation can predict the internal flow characteristics of the slurry pump under different working conditions in advance, providing guidance for test scheme design; model tests provide verification data for simulation results, correct simulation parameters, and improve simulation accuracy. This integration mode is not only applicable to the performance analysis of traditional slurry pumps but also provides an efficient analysis method for the research and development of new slurry pumps, which can significantly shorten the research and development cycle and reduce the research and development cost. Description of the Drawings

[0021] Figure 1It is the working principle diagram of the mud pump performance analysis method based on numerical simulation and model test described in the present invention; Figure 2 It is the design diagram for working condition type determination; Figure 3 It is the design diagram for classification of operating environment data; Figure 4 It is the design diagram for generating working condition optimization strategies; Figure 5 It is the design diagram for the mud pump performance analysis database. Specific implementation manners

[0022] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0023] Please refer to Figures 1-5 A mud pump performance analysis method based on numerical simulation and model test involved in the present invention is specifically implemented as follows: Based on the design parameters (such as impeller diameter, rotational speed, rated flow, etc.) and working conditions (such as medium density, conveying pressure, operating environment temperature, etc.) of the mud pump, a systematic plan for the mud pump performance analysis process is carried out. First, clarify the numerical simulation content and model test content required for the analysis. The numerical simulation content focuses on the simulation of the internal flow field of the mud pump, and the fluid flow states in key components such as the mud pump impeller and volute are modeled through computational fluid dynamics (CFD) software to analyze parameters such as flow velocity distribution, pressure field, and turbulence characteristics; the model test content includes obtaining the measured performance data of the mud pump under different working conditions, such as flow rate, head, efficiency, power consumption, etc., and real-time data is collected by building a physical test bench and configuring sensors (such as electromagnetic flowmeter, pressure transmitter, power meter, etc.).

[0024] According to the design parameters and working conditions of the mud pump, different working condition types of the mud pump are identified and classified. For example, for different flow rate ranges, medium characteristics or operating loads, typical working condition categories are determined to provide a clear target orientation for subsequent simulations and tests.

[0025] Combined with different working condition types, numerical simulation content and model test content of the mud pump, a mud pump performance analysis process set is generated. This process set covers the specific operation steps, data acquisition nodes and analysis methods for simulation and test under each working condition. At the same time, based on the optimal matching principle (such as from low to high according to working condition complexity, giving priority to data dependency, etc.), the execution order of the performance analysis process set is determined to improve the analysis efficiency and data relevance.

[0026] According to the preset analysis period (such as quarterly, annually or a specific project cycle), different working condition types of the slurry pump and the requirements for slurry pump performance indicators (such as industry standards or enterprise internal control indicators), a corresponding list of simulated parameter settings is generated for each working condition. For example, under the large flow rate working condition, set the grid density, turbulence model parameters, boundary conditions (inlet flow rate, outlet pressure) of the CFD simulation, etc.; at the same time, according to the preset test rules (such as the calculation method of the test sample size, repeated test requirements, etc.), working condition types and performance indicator requirements, a list of test plans for each working condition is generated, including test equipment configuration, test point distribution, data acquisition frequency, etc.

[0027] Based on the execution order of the performance analysis process set, the list of simulated parameter settings, the list of test plans and the content of slurry pump performance analysis (such as the flow field analysis target, performance data comparison dimension, etc.) for each working condition are associated and integrated to form an overall plan for slurry pump performance analysis. This plan is presented in the form of a flow chart or Gantt chart, clearly defining the time nodes, responsible entities and data transfer paths of each link to ensure the standardization and traceability of the analysis process.

[0028] Embodiment 1: For the current working condition, it is necessary to determine the performance analysis boundary based on historical performance data, adjacent working condition data and preset working condition division thresholds. Historical performance data includes various parameter records of the current working condition in past operation cycles, such as flow rate, head, efficiency, power, medium temperature, vibration frequency, etc. These data are usually stored in the slurry pump operation monitoring system or historical database and can be retrieved in real time or imported in batches through a data interface. The historical performance data of adjacent working conditions refers to the data of other working conditions that are continuous with the current working condition in terms of working condition parameters (such as flow rate, pressure). For example, when the current working condition is the rated flow rate working condition, the adjacent working conditions can be set as the small flow rate working condition 10% lower than the rated flow rate and the medium-high flow rate working condition 10% higher than the rated flow rate, and its data range covers all operation records with a difference in working condition parameters from the current working condition within the preset range. The preset working condition division threshold is the allowable range of parameter fluctuations preset according to the slurry pump design standard and industry specifications. For example, the flow rate threshold can be set at ±10% of the rated flow rate, and the pressure threshold can be set at ±15% of the rated pressure. The specific values need to be determined in combination with factors such as the slurry pump type and application scenario.

[0029] When determining the performance analysis boundary, first perform a statistical analysis on the historical performance data of the current working condition, calculate descriptive statistics such as the mean, standard deviation, maximum value, and minimum value of key parameters (such as flow rate and head), and identify the data distribution characteristics. Then, compare the historical performance data of adjacent working conditions with the current working condition data to analyze the parameter change trend. For example, if the current working condition is a large flow rate condition, it is necessary to determine whether its flow rate is continuously higher than the preset flow rate threshold or accidentally reaches the threshold within a specific time period. By comparing the parameter differences between the current working condition and adjacent working conditions and combining the preset working condition division threshold, the performance analysis boundary of the current working condition is determined. Specifically, if the mean flow rate of the current working condition is Q0 and the preset flow rate threshold is ±ΔQ, the performance analysis boundary can be defined as the flow rate range [Q0 - ΔQ, Q0 + ΔQ]. All historical data within this range are included in the analysis scope, and data outside this range are considered to belong to adjacent working conditions or abnormal data and need to be processed separately.

[0030] Next, extract the performance analysis data subset of the current working condition according to the performance analysis boundary and the preset time window. The preset time window is to ensure the timeliness and relevance of the analysis data, which can be set according to the operation cycle, maintenance cycle of the mud pump or project requirements, such as the past 3 months, the past 6 months or the last 100 operating hours. The setting of the time window needs to balance the richness and timeliness of the data, avoiding too much old information in the data due to too long a time, or insufficient data volume due to too short a time.

[0031] When extracting the data subset, first screen out the historical data that meets the working condition parameter range according to the performance analysis boundary, and then further filter according to the preset time window, only retaining the data within the time window. For example, if the performance analysis boundary is the flow rate range [Q0 - ΔQ, Q0 + ΔQ] and the preset time window is from January 1, 2025 to March 31, 2025, the data subset is all operation records within this time period and with the flow rate in the range [Q0 - ΔQ, Q0 + ΔQ]. The data subset includes numerical simulation data and model test data. The numerical simulation data includes flow field parameters calculated by CFD software, such as the flow velocity, pressure, turbulence intensity, vorticity, etc. at each calculation node, usually stored in the form of grid data files, data tables or visualization graphics (such as velocity vector diagrams, pressure contour maps); the model test data includes the measured data collected by the physical test bench, such as the flow rate, head, efficiency, power consumption, etc. at different test points, which are collected in real time by sensors and stored as time series data or test reports.

[0032] When extracting data subsets, attention should be paid to the integrity and consistency of the data. For missing data, interpolation methods (such as linear interpolation, polynomial interpolation) or data filling algorithms should be used for completion; for abnormal data (such as jump values caused by sensor failures), data cleaning algorithms (such as moving average filtering, median filtering) should be used for elimination or correction. In addition, the timestamp alignment of numerical simulation data and model test data should be carried out to ensure the time correspondence of the simulation data and test data under the same working conditions, facilitating subsequent comparative analysis and correlation modeling.

[0033] After completing the extraction of the data subset, the content of the numerical simulation and model test related data within the performance analysis data subset should be determined as the slurry pump performance analysis content for the current working condition. Specifically, the numerical simulation analysis content includes: the overall flow characteristics of the internal flow field of the slurry pump, such as whether the velocity distribution at the inlet and outlet of the impeller is uniform, and whether there are vortex or backflow phenomena in the volute; the flow parameters of key components, such as the pressure distribution on the blade surface, the boundary layer thickness, and the prediction of the cavitation occurrence area; the correlation analysis between the flow field parameters and the slurry pump performance indicators, such as the influence of the velocity distribution on the head and the loss of efficiency caused by the turbulence intensity. The model test analysis content includes: the measured performance curves under different test conditions (such as the flow rate - head curve, the flow rate - efficiency curve) to verify the accuracy of the numerical simulation results; the fluctuation range and stability of each performance indicator, such as the maximum deviation of the head and the standard deviation of the efficiency; the comparative analysis of the test data and the design parameters to identify the reasons for performance differences, such as manufacturing errors, installation deviations, or changes in medium characteristics.

[0034] When determining the analysis content, the design objectives and actual application requirements of the slurry pump should be combined to clarify the analysis focus. For example, if the slurry pump frequently experiences a decrease in efficiency during operation, the analysis content should focus on the energy loss areas in the flow field and the efficiency influencing factors in the test data; if there is a problem of excessive vibration, the pulsating pressure distribution in the flow field and the coupling relationship between the vibration frequency and the flow field characteristic frequency in the test data should be analyzed. In addition, a priority ranking mechanism for the analysis content should be established. According to the severity and scope of the problem, it is determined to first analyze the key performance indicators (such as head, efficiency), and then analyze the secondary indicators (such as vibration, noise) to ensure the efficiency and pertinence of the analysis process.

[0035] To ensure the traceability and repeatability of the analysis process, the determination process of the performance analysis boundary, the extraction method of the data subset, the screening basis of the analysis content, etc. should be recorded in detail to form an analysis log. The log content includes: data sources (such as database names, test report numbers), parameter settings (such as preset working condition division thresholds, start and end times of the time window), data processing methods (such as interpolation algorithms, filtering algorithms), analysis content lists, etc. The analysis log can be used as a reference for subsequent performance analysis and can also be used to review and verify the rationality and accuracy of the analysis process.

[0036] In summary, by clarifying the performance analysis boundary of the current working condition, extracting targeted data subsets, and determining specific analysis contents, a solid data foundation and clear analysis direction are provided for the performance analysis of the slurry pump, ensuring that subsequent numerical simulations and model tests can closely focus on the actual requirements of the current working condition, and improving the accuracy and practicality of the analysis results. This process fully considers the spatio-temporal characteristics and physical meanings of the data, screens and processes the data through scientific methods, avoids the interference of irrelevant data, and improves the analysis efficiency and quality.

[0037] Example 2: When the flow rate of the current working condition is greater than the preset flow rate threshold (such as 120% of the rated flow rate), it is directly determined as a large flow rate working condition. This preset flow rate threshold is a critical value preset based on the design characteristics and safe operating range of the slurry pump, and can be determined by referring to the slurry pump design manual or industry standards. If the flow rate of the current working condition does not exceed this threshold, the head and efficiency data of the current working condition need to be classified based on the first preset classification algorithm.

[0038] The first preset classification algorithm uses the K-means clustering algorithm. This algorithm divides data points into different clusters through iterative calculations, maximizing the similarity of data points within the cluster and minimizing the similarity between clusters. Before applying this algorithm, the historical head and efficiency data of the current working condition need to be preprocessed. First, extract the head and efficiency parameters from the historical database and remove obvious outliers (such as extreme values caused by sensor failures). Then, perform normalization processing on the data to eliminate the influence of dimensions. The formula is:

[0039] where \(x\) is the original data and \(x'\) is the normalized data.

[0040] Initialize the parameters of the K-means algorithm, including the number of clusters \(k\) (usually set according to experience or business requirements, such as 3 - 5 clusters), the maximum number of iterations (such as 100 times), and the convergence threshold (such as 0.001). Randomly select \(k\) data points as the initial clustering centers, and calculate the Euclidean distance from each data point to each clustering center:

[0041] where \(x\) is the data point, \(c\) j is the \(j\)-th clustering center, \(c\) ji is the value of the \(i\)-th dimension representing the \(j\)-th clustering center, and \(n\) is the data dimension (here it is 2, that is, head and efficiency).

[0042] Assign each data point to the cluster where the nearest cluster center is located, and recalculate the centroid of each cluster as the new cluster center. Repeat this process until the change in the cluster center is less than the convergence threshold or the maximum number of iterations is reached. Finally, k performance category data are obtained, such as high head and high efficiency category, medium head and medium efficiency category, low head and low efficiency category, etc.

[0043] Calculate the central characteristics of each performance category data, including statistics such as mean, variance, covariance, etc. The mean vector μ reflects the central position of this category of data, and the calculation formula is:

[0044] where N is the number of data points in this category, and x i is the i-th data point.

[0045] The variance matrix ∑ describes the degree of dispersion of the data distribution, and the calculation formula is:

[0046] At the same time, according to the working conditions of the current working condition, combined with the performance potential evaluation model in the mud pump design manual, determine the performance potential level of the current working condition. The working conditions include factors such as medium characteristics (such as viscosity, sand content), operating environment (such as temperature, pressure), and pump speed. The performance potential evaluation model is based on the principles of fluid mechanics and empirical formulas, and comprehensively considers the influence of the above factors on the performance of the mud pump. For example, for the working condition of transporting high-viscosity medium, the model will reduce the theoretical efficiency potential of the pump according to the viscosity correction coefficient; for high-temperature environment, the model will consider the limitation of the liquid vaporization pressure on the cavitation performance of the pump.

[0047] Input the working condition parameters of the current working condition into the performance potential evaluation model, and calculate the performance potential score through a multi-dimensional non-linear mapping relationship. This score is normalized to the 0-100 score interval, and the performance potential level of the current working condition is determined according to the preset level division threshold (such as level 1 potential for 0-30 points, level 2 potential for 31-60 points, and level 3 potential for 61-100 points).

[0048] Subsequently, calculate the matching degree between the central characteristics of each performance category data and the standard characteristics of the preset performance category. The preset performance category standard characteristics are derived from the mud pump design specifications and industry standards. For example, the standard characteristics of the high-efficiency working condition are that the head fluctuation range is ±5% and the efficiency is higher than 90% of the rated value, etc. The matching degree calculation uses the cosine similarity algorithm:

[0049] where A and B are the central characteristic vectors of the performance category data and the preset standard characteristic vectors respectively.

[0050] If the matching degree is greater than a preset matching threshold (such as 0.8), the current working condition is determined as a specific performance working condition, such as an efficient working condition, an inefficient working condition, or a critical working condition. For example, if the cosine similarity between the central feature vector of a certain type of data and the standard feature vector of the efficient working condition is greater than 0.8, the working condition corresponding to this type of data can be determined as an efficient working condition. If the matching degree is not greater than the threshold, it is determined as an ordinary performance working condition.

[0051] For the large-flow working condition, its stability needs to be further analyzed. Collect the time-series data of parameters such as head and power under the large-flow working condition, and calculate the volatility of the parameters:

[0052] where x is the parameter sequence.

[0053] If the volatility exceeds a preset stability threshold (such as the head volatility > 8%), then this large-flow working condition is determined as an unstable working condition, and the pressure pulsation characteristics of the flow field and the force condition of the impeller need to be focused on in the subsequent analysis.

[0054] After determining the type of working condition, establish the mapping relationship between the type of working condition and the key points of analysis. For the efficient working condition, the key points of analysis are the key factors maintaining the current performance, such as the hydrodynamic characteristics of the impeller blades and the flow guiding effect of the volute; for the inefficient working condition, focus on analyzing the main sources of energy loss, such as the velocity circulation distribution at the impeller outlet and the secondary flow phenomenon in the pump; for the critical working condition, it is necessary to evaluate the degree of its proximity to the performance boundary, such as parameters like net positive suction head and surge margin.

[0055] To ensure the reliability of the classification results, a cross-validation method is used to evaluate the K-means clustering results. Randomly divide the historical data into a training set and a test set, execute the clustering algorithm on the training set, and then use the test set to verify the stability of the classification results. Calculate the clustering result consistency indicators under different partitions, such as the Rand Index and the Adjusted Rand Index. The closer the indicator value is to 1, the more reliable the classification result is.

[0056] At the same time, establish a dynamic update mechanism for the type of working condition. As the operation time of the mud pump increases, collect new performance data in real time, and re-execute the clustering analysis and matching degree calculation regularly (such as monthly) to adapt to the working condition drift caused by factors such as pump body wear and medium property changes. During the update process, retain the time stamps of the historical classification results to form a record of the evolution of the type of working condition, providing a basis for equipment maintenance and performance prediction.

[0057] In addition, a visualization tool is developed to assist in the process of determining the working condition type. This tool displays the clustering results, the process of calculating the matching degree, and the final working condition type in a graphical interface, supports users to interactively adjust classification parameters (such as the number of clusters, the matching threshold), and real-time displays the adjusted classification results. Through visualization means, the transparency and interpretability of the process of determining the working condition type are improved.

[0058] In practical applications, the above classification algorithm can be encapsulated as a standardized module and integrated into the mud pump monitoring system. The system collects operation data in real time, automatically calls the classification module to determine the current working condition type, and generates a report containing working condition characteristics, performance potential, and analysis suggestions, providing decision-making support for operators.

[0059] To sum up, in this embodiment, through the combination of methods such as flow threshold judgment, clustering analysis, performance potential evaluation, and feature matching, accurate classification of the mud pump working condition type is achieved. This process fully considers the physical characteristics and data characteristics of the mud pump operation, establishes a complete mapping relationship from the original data to the working condition type, and provides a clear target orientation for subsequent performance analysis and optimization. The introduction of the dynamic update mechanism and the visualization tool further improves the practicability and adaptability of the classification method, ensuring that the working state of the mud pump can be accurately identified under different operation stages and working condition conditions.

[0060] Embodiment 3: For the current working condition, the operation environment data is classified based on the second preset classification algorithm (hierarchical clustering algorithm). The operation environment data includes multi-dimensional parameters such as environmental temperature, humidity, vibration intensity, medium viscosity, sand content, inlet pressure, etc. Each parameter is collected in real time by sensors and stored as time series data. The hierarchical clustering algorithm is divided into agglomerative (bottom-up) and divisive (top-down). Here, the agglomerative algorithm is adopted, and the specific steps are as follows: Similarity calculation: Calculate the similarity between each sample (each sample is a combination of multi-dimensional environmental parameters at a certain moment) in the operation environment data. Considering the difference in parameter dimensions, the data is first standardized, and the Z-score standardization method is adopted:

[0061] where z j is the value of the j-th sample after Z-score standardization, x j is the original data value of the j-th sample, μ j is the mean value of this parameter, and σ j is the standard deviation of this parameter. After standardization, the Euclidean distance between sample i and sample k is calculated as the similarity measure:

[0062] Among them, d(i, k) is the Euclidean distance between sample i and sample k, M is the dimension of environmental parameters (for example, M = 6 corresponds to parameters such as temperature, humidity, and vibration intensity), z i,m and z k,m are the m-th standardized parameter values of samples i and k respectively.

[0063] Sample merging and clustering tree construction: Initially, each sample forms a separate class. Then, each time the two closest classes are merged, the distances between the new class and other classes are calculated (using the group average method, that is, the average of the distances between all samples in the two classes is used as the inter-class distance), until all samples are merged into one large class, forming a clustering tree with a tree structure. The nodes of the clustering tree represent classes, the leaf nodes represent the original samples, and the length of the branches represents the inter-class distance.

[0064] Category truncation: According to the preset number of clustering layers (such as L = 3 layers) or the clustering distance threshold T (such as T = 2.5), appropriate categories are truncated from the clustering tree as the data of each operating environment category. For example, if the preset number of clustering layers is 3 layers, the clustering tree is divided into 3 large classes; if the distance threshold is used, the nodes with an inter-class distance less than T are merged into one class. Each truncated category corresponds to a set of operating data with similar environmental parameter characteristics, such as "high temperature and high humidity class" and "low vibration and stable class".

[0065] For each operating environment category data, analyze the influence degree and change frequency of each environmental factor, and screen the key environmental subsets. The influence degree is determined through sensitivity analysis, and the local regression method (LOESS) is used to calculate the absolute value of the correlation coefficient |r m |, and the correlation coefficient calculation formula is:

[0066] In the formula, z i，m is the m-th standardized environmental parameter value of the i-th sample, y i is the performance index value (such as efficiency) of the i-th sample, and are the means of parameter m and the performance index respectively, and N is the number of samples within the category. Retain the parameters with |r m | ≥ 0.3 as the environmental factors with significant influence.

[0067] The change frequency is determined by calculating the number of fluctuations of the parameter within a unit time. For continuous parameters (such as temperature), the sliding window method is used to count the number of times it crosses the preset fluctuation threshold (such as mean ± 5%); for discrete parameters (such as medium type), the number of changes within the time window is counted. The parameters with a change frequency higher than the average level are screened out and intersected with the significantly influential parameters to form a key environmental subset. For example, if the correlation coefficient of temperature |r| = 0.4 and the change frequency is 2 times per hour (higher than the average of 1 time per hour), it is included in the key environmental subset.

[0068] According to the influence characteristics of the key environmental subset and the mean value of the mud pump performance indicators, an influence label is assigned to each operating environment category data. The influence characteristics include the action direction of parameter changes on performance (such as efficiency decreases when temperature increases) and the degree of action (such as efficiency decreases by 0.5% for every 1°C increase). The mean value of the performance indicator is the average value of the performance indicators of all samples within this category, such as the average head H avg , the average efficiency η avg . The influence label adopts the format of "parameter characteristics + influence description", such as "temperature-sensitive (negatively correlated with efficiency)" and "vibration-stable (head fluctuation < 3%)".

[0069] Subsequently, different strategy generation logics are executed according to the working condition type (efficient working condition, inefficient working condition, critical working condition) of the current working condition: If it is an efficient working condition: Match the influence labels of each operating environment category data with the mud pump performance indicators, and analyze the potential influence path of environmental factors on efficient operation. For example, if the label of a certain category is "sensitive to medium sand content (accelerated head attenuation rate)", it is necessary to evaluate whether the increase in sand content leads to increased impeller wear, thereby affecting the head. By establishing a regression model between environmental parameters and performance indicators, key influencing factors are identified, and targeted optimization strategies are generated, such as adding a medium pretreatment link to reduce the sand content, or selecting wear-resistant materials to manufacture the impeller.

[0070] If it is an inefficient working condition: Improvement and optimization strategies are generated based on preset improvement rules and performance improvement goals. The preset improvement rules include adjustment of impeller geometric parameters (such as increasing the blade outlet angle by 5°), modification of the volute flow path (such as reducing the curvature radius of the diffusion section), optimization of the sealing structure (such as using labyrinth seals instead of packing seals), etc., which are preset based on fluid mechanics theory and engineering experience. The performance improvement goal is determined according to design requirements or industry standards, such as increasing the efficiency to more than 85% and reducing the head fluctuation rate to within 5%. By comparing the performance data of the current working condition with the target value, a suitable combination of improvement rules is selected to form an optimization plan. For example, if the reason for inefficiency is insufficient impeller outlet velocity circulation, rules such as increasing the blade outlet angle and widening the volute flow path can be adopted.

[0071] If it is a critical operating condition: Generate a performance stability strategy based on historical performance fluctuation data and performance stability rules. The historical performance fluctuation data includes the parameter fluctuation amplitude, frequency, and corresponding environmental conditions within a certain period in the past (such as 3 months), and the fluctuation patterns are identified through time series analysis (such as the autoregressive moving average model ARMA). The performance stability rules include the setting of safety protection thresholds (such as triggering an alarm when the inlet pressure is lower than 0.1 MPa) and the control parameter adjustment logic (such as maintaining a stable flow rate through variable frequency speed regulation). For example, if the critical operating condition is the condition close to the minimum value of the net positive suction head (NPSH), based on the flow rate and pressure characteristics when cavitation occurs in the historical data, set to automatically open the reflux valve when the NPSH is lower than 1.2 times the design value to increase the pump inlet flow rate to avoid cavitation.

[0072] When generating the performance analysis process set, it is necessary to transform the above classification, screening, label assignment, and strategy generation processes into standardized operation steps, and clarify the input data, processing methods, and output results of each step. For example, the input of the hierarchical clustering analysis step is the standardized operating environment data, and the output is the clustering category labels and central characteristics; the input of the key environment subset screening step is the category data and performance indicators, and the output is the list of key parameters and impact labels.

[0073] To ensure the executability of the process, develop an automated tool to implement the algorithm integration of each step. The tool has functions such as data import, parameter setting (such as the number of clustering layers, correlation coefficient threshold), and result visualization (such as clustering tree diagrams, impact label cloud diagrams), supports users to adjust analysis parameters according to actual needs, and automatically generates a report containing process steps, data charts, and strategy suggestions.

[0074] In addition, establish a process verification mechanism to evaluate the effectiveness of the process by comparing the historical analysis results and actual performance under different operating condition types. For example, for the high-efficiency operating condition marked as "temperature-sensitive", statistically analyze the actual fluctuation amplitude of the performance indicators when the temperature changes to verify the accuracy of the impact labels; for the generated improvement and optimization strategies, numerically simulate and preview their impacts on the flow field and performance to ensure the rationality of the strategies.

[0075] In summary, in this embodiment, the operating environment data is classified through the hierarchical clustering algorithm, the key factors are screened by combining the influence degree and change frequency, and the analysis process and strategy corresponding to the operating condition type are formed. This process transforms multi-dimensional environmental data into impact labels with physical meanings, realizes the correlation analysis between environmental factors and the performance of the slurry pump, and provides different optimization paths for different operating condition types. The introduction of the automated tool and verification mechanism ensures the standardization of the process and the reliability of the strategies, laying a foundation for the dynamic analysis and continuous optimization of the slurry pump performance.

[0076] Example 4: Taking the performance analysis of a certain type of slurry pump (rated flow rate: 1000 m³ / h, rated head: 50 m) as an example, the preset test rules run through the entire process of the model test, covering key links such as dynamic adjustment of sample size, triggering of supplementary tests, and equipment calibration compensation. The specific applications are as follows: I. Dynamic adjustment of sample size and number of repetitions: In the model test under the large flow rate condition (flow rate: 1200 m³ / h), the sample size is first determined based on historical test data. Assume that the standard deviation of efficiency under this condition in historical data is 2.5%, and the error range under 95% confidence level is set as ±1.5%. Calculate the required sample size according to the statistical formula: at least 11 groups of valid data need to be collected to meet the accuracy requirements. At the initial stage of the test, the test is carried out according to this sample size. In the first 3 groups of tests, the measured efficiency values are 78%, 80%, and 77% respectively, and their average value is 78.3%, and the standard deviation is 1.58%, all within the preset error range. However, the 4th group of data suddenly becomes 72%, exceeding the historical standard deviation range, and is determined as abnormal data, automatically triggering the repeated test mechanism. This condition is repeatedly tested 3 times, and the results are 79%, 78%, and 79% respectively. After merging with the first 3 groups of data and recalculating, the standard deviation drops to 1.1%, meeting the stability requirements. Finally, 6 groups of valid data are retained (the original 3 groups of normal data + 3 groups of repeated test data), and the remaining abnormal data are marked as invalid and excluded.

[0077] In the test under the normal performance condition (flow rate: 800 m³ / h), if the head fluctuation of the first 5 groups of test data is less than 3%, it is determined that the data stability is good and there is no need to increase the number of repetitions; if the head deviation of two consecutive groups of data exceeds 5%, 2 additional repeated tests will be automatically added until the deviation of three consecutive groups of data is less than 3%. This mechanism avoids data distortion caused by accidental interference and ensures the reliability of the test results.

[0078] II. Supplementary test triggering mechanism: In the test under a certain critical operating condition (close to the cavitation boundary, with a flow rate of 900 m³ / h and an inlet pressure of 0.08 MPa), the measured head was 42 m, while the numerical simulation result was 45 m, with a deviation of 6.7% (exceeding the preset error threshold of ±5%). The system automatically triggered the supplementary test module. First, the status of the test equipment was checked: the electromagnetic flowmeter (accuracy class 0.5) showed stable flow, and the pressure transmitter (accuracy class 0.25) was within the calibration validity period. However, slight vibration was found in the inlet pipeline. It was speculated that the vibration might cause pressure measurement errors. Therefore, a vibration damping bracket was installed on the inlet pipeline, and a high-precision pressure sensor (accuracy class 0.1) was replaced for retesting. In the second test, the measured head was 44 m, and the deviation from the simulation result was reduced to 2.2%, meeting the error requirements. During the supplementary test, the boundary conditions of the CFD model were adjusted synchronously. The inlet turbulent intensity was corrected from the preset 5% to 3% to match the actual flow state and avoid deviations caused by the mismatch between the model assumptions and the actual operating conditions.

[0079] If the deviation still exceeds the threshold after the supplementary test, the multi-physical field coupling analysis process is started: compare the pressure distribution in the flow field simulation with the pressure pulsation data measured in the test, check for physical factors that have not been considered (such as medium compressibility, impeller elastic deformation, etc.), and accordingly correct the model parameters or adjust the test plan. For example, if it is found that the simulation does not consider the influence of the sand content in the medium on the turbulent characteristics, the sand content measurement can be added in the test, and the discrete phase model (DPM) can be enabled in the CFD model for recalculation.

[0080] III. Equipment calibration compensation mechanism: Before each test, calibration compensation is carried out on key equipment such as electromagnetic flowmeters, pressure transmitters, and power meters. Taking the calibration of the pressure transmitter as an example: A standard pressure gauge (accuracy class 0.1, range 0 - 1 MPa) is installed in parallel with the test pressure transmitter on the outlet pipeline of the mud pump. Clear water is introduced and the pump speed is adjusted to make the outlet pressure stable at 10%, 50%, and 90% of the range points (i.e., 0.1 MPa, 0.5 MPa, 0.9 MPa) in turn. Record the reading differences between the standard gauge and the test gauge. For example, at 0.5 MPa, the reading of the standard gauge is 0.502 MPa, and the reading of the test gauge is 0.498 MPa, with a deviation of -0.004 MPa. The calibration compensation coefficient is calculated as:

[0081] The original data of the test gauge is corrected by multiplying it with this compensation coefficient. The calibration process for other equipment (such as electromagnetic flowmeters) is similar. By comparing with a standard flowmeter (such as a positive displacement flowmeter), a linear calibration equation is established:

[0082] Among them, a is the slope correction coefficient and b is the intercept correction coefficient, which are determined by regression analysis of at least 3 calibration points.

[0083] During the calibration process, if it is found that the deviation of a certain device exceeds the allowable range of its accuracy class (such as the deviation of the pressure transmitter > ±0.25% FS), it is determined that the device needs to be returned to the factory for repair or replacement, and the test is suspended until the device is calibrated qualified. For example, if the full-scale deviation of the power meter reaches 1.5% during calibration (its accuracy class is 0.5 level), the device is immediately deactivated, and a spare power meter is used for re-calibration to ensure the accuracy of the test data.

[0084] IV. Cross-condition correlation design of the test plan: When generating the test plans for each condition, pay attention to the correlation between conditions and data reusability. For example, in the tests of the large flow rate condition (1200 m³ / h) and the high-efficiency condition (1000 m³ / h, efficiency > 85%), some test equipment and installation processes are shared, but the distribution of test points is adjusted specifically: the large flow rate condition focuses on monitoring the pressure pulsation at the impeller inlet (5 high-frequency pressure sensors are arranged), and the high-efficiency condition focuses on the flow velocity uniformity at the volute outlet (the particle image velocimetry technology PIV is used to measure the flow field). At the same time, expansion interfaces are reserved in the test plan so that some calibration data and installation structures can be reused for subsequent conditions (such as the condition of changing the sand content in the medium), reducing repetitive labor.

[0085] V. Data traceability and process record: During the test process, information such as equipment calibration parameters, reasons for sample size adjustment, and trigger conditions for supplementary tests are recorded in real time to form a complete test log. The log content includes: Calibration time, calibration equipment model, standard instrument number, calibration deviation, and compensation parameters; Initial value, adjusted value, and adjustment basis (such as standard deviation exceeding the limit, abnormal data identification) of the sample size for each condition; Trigger time, cause analysis, adjustment measures, and re-test results of the supplementary test; Auxiliary information such as test personnel and equipment operation status (such as pump speed fluctuation range, medium temperature change).

[0086] For example, in a certain test, the log record shows that: due to the abnormality of the 4th group of data in the large flow rate condition, the repeated test was triggered 3 times. Among them, the 2nd repeated test was invalid due to loose sensor wiring. After re-wiring, the test was successful, and finally the 1st and 3rd repeated data were used. Such detailed records provide a basis for subsequent data review and problem traceability, ensuring the reproducibility of the test process.

[0087] VI. Multidisciplinary collaborative verification: The model test plan and numerical simulation form a closed-loop verification mechanism. Before the test, the positions of the test points and the data acquisition frequency are preset based on the simulation results; during the test, the measured data is fed back to the simulation team in real time to correct the boundary conditions or mesh division; after the test, by comparing the flow field cloud map with the PIV measurement results and the performance curve with the simulation curve, the accuracy of the model is evaluated. For example, in the test under low-efficiency conditions, the measured efficiency is 4% lower than the simulated value. After analysis, it is found that the simulation did not consider the clearance leakage between the impeller and the volute. Therefore, a clearance flow module is added to the model, and after recalculation, the error between the simulation result and the test is reduced to 1.5%.

[0088] Through the implementation of the above preset test rules, the test link of the mud pump performance analysis has achieved the goals of controllable data quality, traceable process, and locatable problems. The dynamic sample size adjustment avoids data redundancy or insufficiency, the supplementary test mechanism ensures the systematic investigation of deviation problems, the equipment calibration compensation guarantees the data accuracy from the source, and the cross-condition correlation design and multi-disciplinary collaboration improve the overall analysis efficiency. The comprehensive application of these rules enables the model test to accurately reflect the actual performance of the mud pump, providing a reliable measured basis for the optimization of numerical simulation and the formulation of performance improvement strategies.

[0089] Example 5: The following takes a mud pump (model NB-1200, rated power 800kW) used in a certain oilfield as an example for detailed description: I. Database architecture design and data storage: The mud pump performance analysis database adopts a hierarchical architecture design, including the raw data layer, cleaning layer, analysis layer, and application layer. The raw data layer stores the unprocessed numerical simulation results (such as the mesh files and pressure distribution data calculated by CFD), the original records of the model test (such as the flow-time series in CSV format collected by sensors), the design parameter documents (such as CAD drawings and technical specifications), and the process data of the working condition type division (such as the intermediate results of cluster analysis). Taking the large-flow working condition (1200m³ / h) as an example, the raw data layer stores the iteration logs of 20 groups of CFD simulations, the original waveform files (sampling frequency 100Hz) of 5 model tests, and the initial center matrix of K-means clustering under this working condition.

[0090] The cleaning layer processes the raw data through the ETL (Extract-Transform-Load) process: removing outliers caused by sensor failures in model tests (such as data points where the head suddenly becomes negative in a certain group of tests), unifying the units of numerical simulation results (such as unifying the pressure unit to MPa), and adding timestamps and operating condition labels to all data (such as "2025-03-15_efficient operating condition"). For example, when processing the operating data in April 2025, the cleaning layer automatically identified and corrected the dimensional error of the medium temperature parameter (mistakenly recording °C as K), and completed the missing flow rate data caused by communication interruption during a certain period through linear interpolation.

[0091] The analysis layer stores the structured data generated through the performance analysis process, including the list of simulation parameter settings for each operating condition (such as the selection of turbulence models and the number of grids in the CFD model), the list of test plans (such as test point coordinates and sensor layout diagrams), the performance analysis report (such as the analysis of the vortex intensity distribution in the flow field and the comparison curve of head efficiency), and the record of the evolution of operating condition types (such as the time node and reason for a certain operating condition being reclassified from "ordinary performance" to "inefficient operating condition"). For example, under the "stability analysis of critical operating conditions" folder in the analysis layer, 8 records of performance stability strategy adjustments during the period from January to May 2025 for this operating condition are stored, including the opening threshold of the reflux valve, the triggering conditions, and the corresponding historical fluctuation data segments for each adjustment.

[0092] The application layer provides functional interfaces such as data query and visualization analysis. Operators can retrieve data through the Web interface according to the time range (such as 2025Q2), operating condition type (such as "efficient operating condition"), or analysis content (such as "vibration analysis"), and view the comparison chart of simulation parameters and test results under a certain operating condition. For example, when querying the data of the efficient operating condition in May 2025, the application layer automatically generates an overlay chart of the flow field pressure contour map and the measured pressure curve for this operating condition, intuitively showing the coincidence degree between the simulation and the test.

[0093] II. Data Cleaning and Maintenance Mechanism: The regular cleaning and maintenance process is automatically started on the 10th of each month. First, duplicate data is identified through the hash value comparison algorithm. For example, during the cleaning in May 2025, it was found that 3 groups of model test data were repeatedly uploaded due to system misoperation. After locating through hash value matching, the redundant records were deleted. For invalid data, the Z-score algorithm is used to detect outliers: setting the threshold as ±3σ, and automatically marking and deleting the data points outside this range (such as records with negative power consumption under a certain operating condition).

[0094] During the maintenance process, it is also necessary to handle the cross-version compatibility issues of data. For example, when the CFD software is upgraded from Ansys 2024 to Ansys 2025, the mesh file format of the old version cannot be directly read. It is necessary to convert it to the new format through the data conversion tool built into the database and record the conversion log for traceability. For design parameter changes (such as the impeller material changing from cast iron to stainless steel), the system automatically generates a data version branch, enabling the new version parameters while retaining the historical version data to ensure the continuity of the analysis results.

[0095] III. Long-term Tracking and Comparative Analysis Process: When conducting the performance trend analysis of the slurry pump based on the database, first generate a time series report of key performance indicators quarterly. Taking the head as an example, the report includes statistics such as the average head of each quarter, the maximum head fluctuation range, and the correlation coefficient between the head and the sand content of the medium. During the tracking from Q1 to Q3 in 2025, it was found that the average head in Q3 decreased by 2.3 m compared to Q1, and the average sand content of the medium increased from 5 kg / m³ to 8 kg / m³ during the same period. It is initially speculated that the decrease in head is related to impeller wear.

[0096] The comparative analysis module supports cross-comparison of data under multiple working conditions and multiple periods. For example, comparing the high-efficiency working condition data in 2025 with the data in the same period in 2024, it is found that the average turbulent intensity at the impeller outlet increased by 15%, while the efficiency decreased by 4%. Combining with the results of the flow field simulation, it is judged that the increase in turbulent intensity led to an increase in energy loss. Such comparisons provide a basis for the equipment maintenance strategy: during the major overhaul in October 2025, apply a wear-resistant coating to the impeller and optimize the surface roughness of the volute flow channel to reduce the turbulent intensity.

[0097] Long-term tracking also includes the evaluation of the performance analysis process itself. For example, by comparing the results of the working condition type division in different periods, verify the stability of the clustering algorithm: in May 2025, the system automatically detected that the deviation amplitude of the clustering center of a certain working condition exceeded the preset threshold (10%), prompting the need to retrain the clustering model to avoid classification deviations caused by changes in data distribution.

[0098] IV. Data Security and Permission Management: The database adopts role-based access control (RBAC), dividing into three types of roles: "administrator", "analyst", and "operator". The administrator has the highest permissions such as data deletion and schema modification; the analyst can create analysis tasks and call algorithm interfaces; the operator can only view the real-time data related to the current work (such as the test results of the current operating conditions). For example, the operator can only access the real-time collected flow rate and pressure data on the monitoring interface and cannot view the details of the simulation parameter settings of historical working conditions.

[0099] Data transmission and storage are encrypted: data is transmitted via the HTTPS protocol, and the storage layer uses the AES-256 encryption algorithm to encrypt sensitive information (such as design parameters and optimization strategies). Data is backed up regularly, and the backup files are stored on an off-site server to ensure that the system can be restored to the most recent valid state (such as a complete data snapshot on April 30, 2025) in the event of a system failure or data loss.

[0100] 5. Examples of actual application scenarios: In a shale gas extraction project, the mud pump needs to operate for a long time in a high-sand medium (sand content 10-15kg / m³) environment. Through database tracking, it was found that the mud pump efficiency under this condition decreased at a rate of 0.8% per month, and the erosion wear depth at the impeller inlet increased by 0.2mm every quarter. Based on these data, the engineering team adjusted the maintenance cycle: the frequency of impeller inspection was shortened from once every 6 months to once every 3 months, and wear-resistant impeller spare parts were reserved in advance. At the same time, by comparing the flow field simulation results of different sand content conditions, the diversion structure of the suction pipe was optimized, the impact angle of the medium on the impeller was reduced, and the wear process was delayed.

[0101] In another scenario, the mud pump frequently had vibration over-limit alarms in the high temperature environment in summer. By retrieving historical data from the database, it was found that the vibration intensity was positively correlated with the ambient temperature (correlation coefficient 0.78), and the operating environment category label during the high temperature period was "temperature sensitive (intensified vibration)". Based on this, the team added the start and stop threshold adjustment function of the temperature-controlled fan in the cooling system. When the ambient temperature exceeded 35°C, auxiliary cooling was automatically started, the bearing temperature was controlled within 70°C, and the vibration intensity was reduced to below the alarm threshold.

[0102] 6. Data-driven decision support: The database provides support for the full life cycle management of mud pumps: during the design phase, the number of impeller blades is optimized by referring to the flow field analysis data of historical operating conditions (for example, increasing from 5 to 6 to improve the efficiency under low flow conditions); during the manufacturing phase, quality control standards for key components are established based on test data (such as the impeller casting tolerance of ±0.05mm); during the operation and maintenance phase, the remaining life of the equipment is predicted through performance trend analysis (for example, the impeller replacement cycle is calculated to be 18 months based on the wear rate).

[0103] For example, the database analysis in June 2025 showed that the frequency of critical operating conditions of a mud pump increased by 30% compared with the previous year, and the number of triggering of the performance stability strategy increased by 50% year-on-year. After evaluation, it was determined that the pump had entered the aging stage, and it was recommended to include it in the annual upgrade and renovation plan, replace it with an energy-efficient impeller and install intelligent monitoring sensors to improve operational stability.

[0104] In this embodiment, through the construction of a structured database, the formulation of a standardized cleaning process, and the implementation of multi-dimensional tracking and analysis, the full-life cycle management of the performance data of the mud pump is achieved. This mechanism not only provides reliable data support for current performance analysis, but also reveals the internal laws of the performance evolution of the mud pump through in-depth mining of historical data, providing a scientific basis for equipment maintenance, design optimization, and operation decision-making. Data security and permission management ensure the reasonable use of information, and cross-scenario application cases demonstrate the effectiveness and practicality of this mechanism in actual projects.

[0105] It should be noted that in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise", or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or device.

[0106] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for analyzing the performance of a mud pump based on numerical simulation and model tests, characterized in that, The method includes: Planning the performance analysis process of the slurry pump based on the design parameters and working conditions of the slurry pump, determining the numerical simulation content and model test content required for the analysis. The numerical simulation content includes the internal flow field simulation of the slurry pump, and the model test content includes obtaining the measured performance data of the slurry pump under different working conditions; Determining different working condition types of the slurry pump according to the design parameters and working conditions of the slurry pump; Generating a slurry pump performance analysis process set according to the different working condition types, numerical simulation content and model test content of the slurry pump, and determining the execution order of the performance analysis process set based on the optimal matching principle; Generating a simulation parameter setting list for each working condition according to the preset analysis period, different working condition types of the slurry pump and the slurry pump performance index requirements, and generating a test plan list for each working condition according to the preset test rules, different working condition types of the slurry pump and the slurry pump performance index requirements; Associating the simulation parameter setting list, test plan list and slurry pump performance analysis content of each working condition based on the execution order of the performance analysis process set to generate the overall plan for the slurry pump performance analysis.

2. The method for analyzing the performance of a slurry pump based on numerical simulation and model test according to claim 1, wherein Planning the performance analysis process of the slurry pump based on the design parameters and working conditions of the slurry pump, determining the numerical simulation and model test content required for the analysis, including: For the current working condition, determining the performance analysis boundary of the current working condition according to the historical performance data of the current working condition, the historical performance data of adjacent working conditions and the preset working condition division threshold; Extracting the performance analysis data subset of the current working condition according to the performance analysis boundary and the preset time window; Determining the content of the numerical simulation and model test related data within the performance analysis data subset as the slurry pump performance analysis content of the current working condition.

3. The method for analyzing the performance of a slurry pump based on numerical simulation and model test according to claim 1, characterized in that, Determining different working condition types of the slurry pump according to the design parameters and working conditions of the slurry pump, including: For the current working condition, when the flow rate of the current working condition is greater than the preset flow rate threshold, determining that the current working condition is a large flow rate working condition; When the flow rate of the current working condition is not greater than the preset flow rate threshold, classifying the head and efficiency data of the current working condition based on the first preset classification algorithm to obtain each performance category data, and determining the central characteristics of each performance category data; Determining the performance potential level of the current working condition according to the working conditions of the current working condition; When the matching degree between the central characteristics of each performance category data and the preset performance category is greater than the preset matching threshold, determining that the current working condition is a specific performance working condition; When the matching degree is not greater than the preset matching threshold, determining that the current working condition is a general performance working condition.

4. The method for analyzing the performance of a mud pump based on numerical simulation and model tests according to claim 3, wherein The slurry pump performance analysis content further includes slurry pump structure data and operating environment data. The different working condition types of the slurry pump include high-efficiency working conditions, low-efficiency working conditions and critical working conditions. Generating a slurry pump performance analysis process set according to the different working condition types, numerical simulation content and model test content of the slurry pump, including: For the current working condition, classify the operation environment data of the current working condition based on the second preset classification algorithm to obtain data of each operation environment category; for the current operation environment category data, filter the current operation environment category data based on the influence degree and change frequency of each environmental factor in the current operation environment category data to obtain a key environment subset of the current operation environment category data, and determine an influence label of the current operation environment category data according to the influence characteristics of the key environment subset and the mean value of the slurry pump performance index. According to whether the working condition type of the current working condition is an efficient working condition.

5. The method for analyzing the performance of a slurry pump based on numerical simulation and model tests according to claim 4, characterized in that, After determining whether the working condition type of the current working condition is an efficient working condition, it further includes: If the working condition type of the current working condition is an efficient working condition, match the influence labels of the data of each operation environment category with the slurry pump performance index to generate a targeted optimization strategy. If the working condition type of the current working condition is an inefficient working condition, generate an improvement optimization strategy based on a preset improvement rule and a performance improvement target. If the working condition type of the current working condition is a critical working condition, generate a performance stability strategy according to historical performance fluctuations and performance stability rules.

6. The method for analyzing the performance of a mud pump based on numerical simulation and model test according to claim 5, wherein After generating the slurry pump performance analysis process set, it further includes: Update the slurry pump performance analysis content of each working condition based on the real-time operation data of the slurry pump. Recalculate the working condition type of each working condition according to the updated slurry pump performance analysis content. Dynamically adjust the execution order of the performance analysis process set according to the recalculated working condition type.

7. The method for analyzing the performance of a slurry pump based on numerical simulation and model test according to claim 1, wherein The preset test rule includes: Dynamically adjust the sample size and repetition times of the model test based on the confidence interval of the slurry pump historical test data. When the deviation between the measured performance data and the numerical simulation result exceeds a preset error threshold, automatically trigger the supplementary test module.

8. The method for analyzing the performance of a slurry pump based on numerical simulation and model tests according to claim 7, characterized in that The preset test rule further includes: Establish a calibration compensation mechanism for the test equipment, and perform error compensation calibration on the test equipment based on the reference values of the standard flowmeter and pressure gauge before each test.

9. The method for analyzing the performance of a slurry pump based on numerical simulation and model tests according to claim 1, characterized in that After generating the overall scheme of the slurry pump performance analysis, it further includes: Establish a slurry pump performance analysis database, and store the simulation parameter setting list, test scheme list, slurry pump performance analysis content, and data of each working condition type in the overall scheme into this database. Regularly clean and maintain the data in the database, and remove invalid data and duplicate data. Based on this database, conduct long-term tracking and comparative analysis on the slurry pump performance analysis results to evaluate the change trend of the slurry pump performance over time.

10. The method for analyzing the performance of a slurry pump based on numerical simulation and model tests according to claim 4, characterized in that, The second preset classification algorithm is a hierarchical clustering algorithm. The specific steps for classifying the operation environment data of the current working condition include: Calculate the similarity between each sample in the operation environment data. Gradually merge the samples into different categories according to the similarity to form a clustering tree. According to the preset number of clustering layers or clustering distance threshold, intercept appropriate categories from the clustering tree as data of each operation environment category.

Images