Intelligent cleaning control method and system for emulsion pump

Abstract

The present invention discloses an intelligent cleaning control method and system for an emulsion pump, which relates to the field of intelligent cleaning technology. The method comprises: calibrating the relative state curve of the emulsion pump, extracting data from a cleaning record database, and clustering and resampling the data to generate k working condition reorganization sample sets; training k pollution rate mapping models based on the reorganized sample data and the relative pump state curve to predict the cumulative pollution state of the pump; monitoring the operating parameters of the pump in real time to evaluate the pollution state, integrating the cumulative pollution state with the monitored pump state, and making cleaning decisions and cleaning control. The present invention solves the technical problem that existing emulsion pump cleaning methods lack accurate pollution state assessment and dynamic adaptability, resulting in excessive or insufficient cleaning operations. The present invention achieves the technical effect of dynamically adjusting the cleaning operation through real-time monitoring and pollution state assessment, realizing accurate cleaning control, and avoiding excessive or insufficient cleaning operations.

Description

Technical Field

[0001] The present invention relates to the field of intelligent cleaning technology, and in particular to an intelligent cleaning control method and system for an emulsion pump. Background Art

[0002] Emulsion pumps, a common piece of equipment in industrial production, are widely used in various fluid conveying and circulation systems. However, over extended operation, their performance can gradually decline due to contamination, impacting their efficiency and service life. Traditional cleaning methods often rely on fixed cleaning cycles or manual judgment, making it difficult to accurately assess the pump's contamination status. This can lead to over- or under-cleaning, impacting production efficiency and equipment maintenance costs. Summary of the Invention

[0003] The present application provides an intelligent cleaning control method and system for an emulsion pump, which is used to solve the technical problem that existing emulsion pump cleaning methods lack accurate pollution status assessment and dynamic adaptability, resulting in excessive or insufficient cleaning operations.

[0004] The first aspect of the present application provides an intelligent cleaning control method for an emulsion pump, the method comprising: combining target scene requirements with a target emulsion pump, calibrating a relative pump state curve, wherein the relative pump state represents the pollution progress of the target emulsion pump; obtaining unique identification information of the target emulsion pump, and extracting data from a cleaning record database based on the unique identification information; clustering and resampling the data extraction results to obtain reconstructed sample data, wherein the data extraction results include sample pump control parameters and sample pump state parameters, and the reconstructed sample data includes k working condition reconstructed samples set; based on the recombined sample data and the relative pump state curve, construct and train k relative pollution rate mapping models, wherein the relative pollution rate mapping model takes the pump control parameter as input; extract the current cycle operation log of the target emulsion pump, and predict the cumulative pump state of the target emulsion pump based on the current cycle operation log and the k relative pollution rate mapping models; obtain the pump operation parameter information in real time, and perform direct pollution state evaluation based on the relative pump state curve to obtain the monitoring pump state; fuse the cumulative pump state and the monitoring pump state to make cleaning decisions and clean control.

[0005] The second aspect of the present application provides an intelligent cleaning control system for an emulsion pump, the system comprising: a relative pump state curve calibration module, the relative pump state curve calibration module being used to calibrate the relative pump state curve in combination with the target scenario requirements and the target emulsion pump, wherein the relative pump state represents the pollution progress of the target emulsion pump; a data extraction module, the data extraction module being used to obtain the unique identification information of the target emulsion pump, and extracting data from the cleaning record database based on the unique identification information; a clustering segmentation and resampling module, the clustering segmentation and resampling module being used to cluster and resample the data extraction results to obtain recombined sample data, wherein the data extraction results include sample pump control parameters and sample pump state parameters, and the recombined sample data includes k working condition recombined sample sets; a rate mapping model construction module The rate mapping model construction module is used to construct and train k relative pollution rate mapping models based on the recombined sample data and the relative pump state curve, wherein the relative pollution rate mapping model takes the pump control parameters as input; the cumulative pump state prediction module, the cumulative pump state prediction module is used to extract the current cycle operation log of the target emulsion pump, and predict the cumulative pump state of the target emulsion pump based on the current cycle operation log and the k relative pollution rate mapping models; the direct pollution state evaluation module, the direct pollution state evaluation module is used to obtain the pump operation parameter information in real time, and perform direct pollution state evaluation based on the relative pump state curve to obtain the monitoring pump state; the cleaning decision control module, the cleaning decision control module is used to integrate the cumulative pump state and the monitoring pump state to perform cleaning decision and cleaning control.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The present application provides an intelligent cleaning control method and system for an emulsion pump, which relates to the field of intelligent cleaning technology. By calibrating the relative pump state curve, the cleaning record data is extracted in combination with unique identification information, and the data is clustered and resampled. The pollution rate mapping model is trained to predict the cumulative pollution state of the pump, the pump operating parameters are monitored in real time, and the pollution is evaluated based on the state curve. The accumulation and monitoring states are integrated to achieve accurate cleaning decision-making and control. This solves the technical problem that the existing emulsion pump cleaning methods lack accurate pollution state evaluation and dynamic adaptability, resulting in excessive or insufficient cleaning operations. It achieves the technical effect of dynamically adjusting the cleaning operation through real-time monitoring and pollution state evaluation, realizing accurate cleaning control, avoiding excessive or insufficient cleaning operations, and thereby improving the operating efficiency of the pump and extending the life of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0009] Figure 1 A schematic flow chart of an intelligent cleaning control method for an emulsion pump provided in an embodiment of the present application;

[0010] Figure 2 A schematic structural diagram of an intelligent cleaning control system for an emulsion pump provided in an embodiment of the present application.

[0011] Explanation of the accompanying symbols: relative pump state curve calibration module 11, data extraction module 12, cluster segmentation and resampling module 13, rate mapping model construction module 14, cumulative pump state prediction module 15, direct pollution state evaluation module 16, cleaning decision control module 17. DETAILED DESCRIPTION

[0012] The present application provides an intelligent cleaning control method and system for an emulsion pump, which is used to solve the technical problem that existing emulsion pump cleaning methods lack accurate pollution status assessment and dynamic adaptability, resulting in excessive or insufficient cleaning operations.

[0013] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only some of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0014] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices.

[0015] Example 1, as Figure 1 As shown, the present application provides an intelligent cleaning control method for an emulsion pump, the method comprising:

[0016] P10: Based on the target scenario requirements and the target emulsion pump, calibrate the relative pump status curve, where the relative pump status represents the pollution progress of the target emulsion pump.

[0017] Furthermore, step P10 in the embodiment of the present application further includes:

[0018] P11: Interact with the target emulsion pump to obtain the key status indicator set of the pump; P12: Based on the target scenario requirements, determine the evaluation weights and performance control limits of the key status indicator set; P13: Based on the evaluation weights and the key status indicator set, define a relative performance evaluation function, and calibrate the relative pump status curve of the target emulsion pump in combination with the performance control limits.

[0019] It should be understood that calibrating the relative pump state curve, combining the target scenario requirements with the actual conditions of the target emulsion pump, is a key step in ensuring the emulsion pump's stable operation under varying contamination conditions. The relative pump state represents the contamination progression of the emulsion pump, reflecting changes in pump performance and efficiency through a series of key state indicators.

[0020] Specifically, the system first interacts with the target emulsion pump to obtain a set of key status indicators. These indicators are crucial for evaluating the pump's performance and contamination level. These indicators may include flow rate, pressure, energy efficiency ratio, and temperature fluctuations, all of which directly impact the pump's operating efficiency and contamination status. By monitoring these key indicators in real time, it is possible to reflect changes in the pump's operation as the contamination progresses.

[0021] Next, these key health indicators are weighted and evaluated based on the specific application scenario and requirements, and performance control limits (PCLs) are determined for each indicator. PCLs represent the maximum degree of performance degradation that the pump can tolerate during operation. For example, if the ideal performance value for a particular indicator is 100%, the PCL might be 90%, indicating a 10% tolerance for performance degradation. This step ensures that the acceptable performance degradation range for the pump in the event of contamination is clearly defined, allowing for assessment of the pump's health.

[0022] Furthermore, after obtaining the weights and control limits for each indicator, the next step is to define a relative performance evaluation function. This function converts the performance data of multiple indicators into a comprehensive evaluation value. The value of this function is not strictly required to be between 0 and 1. Its purpose is to convert the multiple performance aspects of the pump into a single evaluation value, facilitating a comprehensive assessment of the pump's overall condition. For example, when all key performance indicators of the pump reach ideal levels, the function value is 1; when the pump's condition is close to the worst, the function value may be close to 0. In this way, the pump's contamination process can be reflected through a comprehensive function value.

[0023] Combining the performance evaluation function and control limits, a relative pump condition curve can be calibrated for the target emulsion pump. This curve, similar to a battery's SOC (State of Charge) curve, shows how the pump's contamination progression correlates with changes in its overall performance. The 100% point on the curve represents ideal pump performance, while the 0% point represents 90% of ideal performance (based on the performance control limits). This calibration helps accurately assess pump performance at different stages of contamination and determine when cleaning is appropriate.

[0024] Through this calibration method, the relationship between the contamination process of the emulsion pump and the pump performance is quantified and visualized, providing important data support and decision-making basis for subsequent intelligent cleaning control.

[0025] P20: Obtain unique identification information of the target emulsion pump, and extract data from the cleaning record database based on the unique identification information.

[0026] Optionally, by obtaining the unique identification information of the emulsion pump, data can be further extracted from the cleaning record database to provide data support for subsequent intelligent cleaning control. This process involves equipment identification and data management, which is the basis for the precise operation of the intelligent cleaning control system.

[0027] First, the target emulsion pump is identified using the system or device's unique identification information. Typically, emulsion pumps are assigned a unique identification code (such as a QR code, RFID tag, or serial number) that uniquely identifies the pump within the device and system. In practical applications, this unique identification information is used to accurately distinguish between different emulsion pumps, preventing the system from mistakenly attributing data to the wrong device. Therefore, unique identification information ensures the accuracy of operations and data analysis.

[0028] This unique identification information is used to access historical cleaning records and operational data related to the pump. Specifically, the system uses the unique identification information to locate the relevant cleaning records for the target pump from the dataset stored in the cleaning records database. This data typically includes the pump's historical cleaning time, frequency, cleaning methods, and post-cleaning performance evaluations. This historical record provides valuable information for determining the current contamination status of the pump and predicting future cleaning needs.

[0029] The cleaning records database is a comprehensive management system that typically contains a large amount of historical operational data. Effective database management and retrieval enable accurate data extraction. The system's database management capabilities are crucial in this step, supporting efficient data access and processing. Generally speaking, data in the database can be accurately extracted using methods such as the SQL query language, ensuring a fast response.

[0030] After data extraction, the extracted historical cleaning records are further analyzed. This data serves as the basis for subsequent analysis, helping the system better understand the pump's cleaning cycles and contamination progression. For example, by analyzing historical cleaning records, the system can determine whether the target pump's cleaning intervals are normal and whether the pump's contamination level has reached the standard requiring cleaning. Based on this data, the system can then determine whether the next cleaning operation should be initiated.

[0031] This step enables accurate data acquisition of the target emulsion pump, providing powerful data support for addressing contamination issues in complex environments. This precise data extraction mechanism not only enhances system reliability but also makes pump management more intelligent and automated.

[0032] P30: Clustering and resampling the data extraction results to obtain reorganized sample data, wherein the data extraction results include sample pump control parameters and sample pump state parameters, and the reorganized sample data includes k working condition reorganized sample sets.

[0033] Furthermore, step P30 in the embodiment of the present application further includes:

[0034] P31: Determine the number of clusters k based on the elbow method, and perform density cluster analysis on the data extraction results based on the sample pump control parameters; P32: Define k standard operating condition intervals based on the k clusters in the cluster analysis results; P33: Use the k standard operating condition intervals as segmentation constraints to segment the data extraction results, and output the segmentation results as k recombined data sets; P34: Based on preset confidence constraints, perform confidence screening on the k recombined data sets to obtain k working condition recombined sample sets.

[0035] Specifically, the data extraction results are clustered and resampled to obtain reorganized sample data, providing strong support for subsequent cleaning control.

[0036] First, based on the data extraction results, the elbow method is used to determine the number of clusters, k. This method is a common clustering algorithm optimization technique that primarily determines the optimal number of clusters by plotting the relationship between the number of clusters, k, and the sum of squared clustering errors for different values. In this step, the system uses the elbow method to analyze sample pump control parameters (such as flow rate, pressure, and energy efficiency ratio) to determine the most appropriate number of clusters. This method effectively separates data samples with similar characteristics, providing a clear classification basis for subsequent data processing and model training.

[0037] After determining the number of clusters, k, the cluster analysis results are used to divide the clusters into k clusters, and k standard operating ranges are further defined. These standard operating ranges are based on the sample characteristics within each cluster, defining a range of operating conditions that matches the characteristics of the samples in that cluster. These operating ranges reflect the pump's performance under different operating conditions, enabling the system to provide more detailed monitoring and control of the pump's status during future operations. For example, if a cluster represents a high-load operating state for the pump, the operating range corresponding to that cluster will be defined as the pump's high-load operating range.

[0038] Next, the data extraction results are segmented using k standard operating condition intervals as segmentation constraints. Specifically, the system segments the extracted data according to the standard operating condition intervals, creating a separate reconstructed dataset for each interval. This allows the system to perform targeted analysis and cleanliness control decisions based on the pump's various operating conditions. Each reconstructed dataset represents the pump's state under a specific operating condition, facilitating subsequent contamination status assessment and model training.

[0039] Finally, the k reconstructed datasets are screened based on preset confidence constraints (such as confidence intervals). Confidence screening involves selecting reliable data samples based on certain statistical criteria (such as a 95% confidence level), removing noise and unreliable data, and thus improving the accuracy of model predictions. This step ensures that the final selected dataset is more accurate and stable, generating k sets of reconstructed operating condition samples. These reconstructed operating condition sample sets will serve as the basis for training the pollution rate mapping model in subsequent steps.

[0040] Through the above processing steps, the raw extracted data is transformed into more meaningful reconstructed sample data based on operating conditions through clustering, segmentation, and screening. This data provides an accurate basis for subsequent contamination status assessments and cleaning decisions, ensuring that the cleaning control of the emulsion pump can be dynamically adjusted according to actual operating conditions, thereby achieving the goal of optimizing pump performance and extending pump life.

[0041] P40: Based on the reorganized sample data and the relative pump state curve, k relative pollution rate mapping models are constructed and trained, wherein the relative pollution rate mapping model uses pump control parameters as input.

[0042] Furthermore, step P40 in this embodiment of the present application further includes:

[0043] P41: Sample pump state parameters based on the reorganized sample data; P42: Combining the relative pump state curve with the sample pump state parameters, the reorganized sample data is relatively state marked to obtain marked sample data; P43: With the relative state marking result as supervision, k relative pollution rate mapping models are constructed and trained based on the marked sample data, wherein the k relative pollution rate mapping models correspond one-to-one to the k standard operating conditions.

[0044] It should be understood that based on the reorganized sample data and the relative pump status curves, k relative contamination rate mapping models are constructed and trained. These models, by inputting pump control parameters, can predict the pump contamination rate and provide real-time contamination status assessment, providing accurate decision support for intelligent cleaning control.

[0045] First, the system utilizes sample pump status parameters from the reconstructed sample data. These parameters typically include key performance indicators such as flow rate, pressure, and energy efficiency. These status parameters are fundamental information for evaluating the pump's operating status and reflect its current operating conditions. By analyzing these sample pump status parameters, the system can identify pump performance under different operating conditions, providing data support for the subsequent construction of a pollution rate model.

[0046] Next, the reconstructed sample data is labeled by combining the relative pump state curve with the sample pump state parameters. The relative pump state curve reflects the changes in pump performance during the contamination process. Using this curve, the system can associate each sample pump state with its corresponding contamination level. Specifically, the system calibrates the contamination progress of each sample based on the pump state parameters and the relative pump state curve, assigning each sample a label value. These label values ​​represent the contamination state of the pump at a specific point in time, thus forming labeled sample data. Labeled sample data provides supervised examples for subsequent model training, improving model accuracy and reliability.

[0047] Next, based on the labeled sample data, the system constructs and trains k relative contamination rate mapping models, using the relative state labeling results as supervision. Each mapping model uses the pump's control parameters (such as flow rate and pressure) as input to predict the pump's contamination rate under different operating conditions. The k mapping models correspond to k standard operating ranges, each representing the pump's performance under different loads or operating conditions. By inputting the labeled sample data into these models, the system trains a set of models that reflect the relationship between the pump's contamination rate and the control parameters. These models will be used to evaluate the pump's contamination rate in real time in real-world applications, enabling the system to effectively control cleaning.

[0048] Supervised learning methods can be used to construct relative contamination rate mapping models. By leveraging labeled sample data, these models can learn the mapping relationship between pump control parameters and contamination rates. Regression algorithms (such as linear regression, decision tree regression, and random forest regression) are used to build these models. Training data is used to continuously optimize parameters and improve prediction accuracy.

[0049] In addition, the definition of k standard operating condition intervals used in the model training process ensures that each mapping model focuses on the pollution rate prediction under specific operating conditions. Through this operating condition differentiation, the system can provide more accurate pollution rate assessment for different operating modes.

[0050] Improve the accuracy of pump contamination assessment and provide an important basis for the dynamic adjustment of intelligent cleaning systems.

[0051] P50: extracting the current cycle operation log of the target emulsion pump, and predicting the cumulative pump state of the target emulsion pump based on the current cycle operation log and the k relative pollution rate mapping models.

[0052] Furthermore, step P50 in the embodiment of the present application further includes:

[0053] P51: Obtain the target pump's current cycle operation log, and extract the operating condition setting sequence from the current cycle operation log; P52: Map and transform the operating condition setting sequence based on k relative pollution rate mapping models to obtain a relative pollution rate sequence; P53: Perform cumulative analysis on the relative pollution rate sequence to obtain the cumulative pump status.

[0054] Optionally, the current cycle operation logs of the target emulsion pump are extracted, and the cumulative pump state of the target emulsion pump is predicted by using these logs and the k trained relative pollution rate mapping models.

[0055] First, the target emulsion pump's current cycle operation logs are obtained. These logs record the pump's operating parameters over a specific period of time, including flow rate, pressure, energy efficiency ratio, temperature, and other information. By extracting the operating condition sequence from these logs, the system can understand the pump's specific operating status during the current cycle. These operating condition sequence reflects the pump's specific operating mode during different time periods. For example, some time periods may indicate high-load operation, while other time periods may indicate low-load or standby mode. Extracting these operating condition sequences facilitates a more detailed analysis of the pump's operation.

[0056] Next, the extracted operating condition setting sequences are input into k relative contamination rate mapping models for mapping conversion. Each model corresponds to a specific operating condition interval. In this way, the system can convert each operating condition into a corresponding relative contamination rate. The relative contamination rate sequence represents the speed or progress of pump contamination under different operating conditions. The mapping process essentially links the pump's operating parameters with the contamination process, reflecting the changes in pump contamination under different operating conditions. These relative contamination rate sequences provide the basis for further cumulative analysis.

[0057] Finally, a cumulative analysis of the relative contamination rate sequence is performed to calculate the contamination progression of the target emulsion pump throughout its entire cycle. The cumulative analysis process chronologically sums the contamination rates for each period to determine the pump's overall contamination status. The key to this process is accurately assessing the progression of contamination throughout the pump's operating cycle. Cumulative analysis provides a more comprehensive understanding of the pump's cumulative contamination. The system generates a cumulative pump status based on these cumulative results, reflecting the pump's degree of contamination throughout its entire cycle and providing a reliable basis for subsequent cleaning control decisions.

[0058] Furthermore, implementing cumulative analysis requires a weighted accumulation mechanism to account for the impact of different time periods on the pump's contamination status. For example, prolonged high-load operation may lead to an accelerated contamination rate, making such a period of time contribute more significantly to the cumulative contamination status. Therefore, cumulative analysis requires more than a simple summation; it may also involve weighting the contamination rates under different operating conditions.

[0059] Through the above steps, using the target emulsion pump's current cycle operation log and relative contamination rate mapping model, the pump's cumulative pump status can be accurately predicted, providing the intelligent cleaning control system with a contamination status assessment based on actual operation data. This helps the system identify contamination problems in a timely manner and take corresponding cleaning measures, thereby improving the pump's operating efficiency and extending the equipment's service life.

[0060] P60: Acquire the operating parameter information of the pump in real time, and perform direct pollution status evaluation based on the relative pump status curve to obtain the monitoring pump status.

[0061] Specifically, the pump status is determined by acquiring real-time pump operating parameter information and combining it with the relative pump status curve to perform a real-time assessment of the contamination status. This process enables real-time monitoring of the pump's operation, promptly reflecting the progress of the pump's contamination, and providing data support for cleaning decisions.

[0062] First, the target emulsion pump's operating parameters, such as flow rate, pressure, temperature, and energy efficiency ratio, are monitored in real time. These operating parameters are crucial for assessing the pump's current operating status. For example, flow rate and pressure reflect the pump's workload, while temperature and energy efficiency ratio indicate its operating efficiency and potential abnormalities. By collecting this data in real time, the system can dynamically track the pump's status, providing timely fundamental data for subsequent contamination assessments.

[0063] After acquiring real-time operating parameters, they are combined with the relative pump status curve to perform a direct contamination assessment. The relative pump status curve is a calibrated curve that reflects the relationship between pump operating parameters (such as flow rate and pressure) and the progress of pump contamination. By inputting real-time operating parameters into the relative pump status curve, the system can assess the pump's contamination status in real time. For example, the pump's flow rate and pressure parameters may change during the contamination process, and this curve can be used to map the current degree of contamination.

[0064] The core of this evaluation method is to quickly assess the pump's contamination status by matching real-time operating data with a calibrated status curve. This method does not rely on excessive historical data, but instead performs contamination detection based on the current real-time operating status, enabling timely response to pump contamination conditions.

[0065] After the contamination status assessment, the monitoring pump status is generated, for example, expressed as a numerical value for the degree of contamination. This monitoring pump status can be expressed as a quantitative indicator, such as a percentage of contamination progress or a relative performance degradation value. These monitoring results reflect the current contamination progress of the pump and provide a basis for subsequent cleaning control. If the monitored pump status reaches a certain contamination threshold, a cleaning procedure can be automatically initiated to ensure the pump operates in ideal working conditions.

[0066] P70: Integrate the accumulated pump status and the monitoring pump status to make cleaning decisions and perform cleaning control.

[0067] For example, cleaning decisions and control are made by integrating accumulated pump status with monitored pump status. The goal is to integrate the pump's contamination process and current contamination status to make more accurate and effective cleaning decisions, thereby ensuring continuous optimization of emulsion pump performance and long-term stable operation of the equipment.

[0068] First, the cumulative pump status and the monitored pump status are integrated. The cumulative pump status is derived by analyzing the pump's operating data over a period of time, reflecting the pump's pollution progression throughout its entire lifecycle. The monitored pump status, on the other hand, assesses the pollution status using a relative pump status curve based on real-time pump operating parameters. The integration of the two aims to combine the pump's historical pollution progression with its current pollution level, forming a comprehensive pollution status assessment.

[0069] Specifically, the Cumulative Pump Status provides a snapshot of the pump's contamination trends over its entire lifecycle, while the Monitored Pump Status reflects the pump's current level of contamination. By combining these two metrics, the system can factor in both long-term contamination accumulation and current contamination rates, providing more comprehensive information for cleaning decisions. For example, if the Cumulative Contamination Status is high but the Monitored Status is low, this may mean that the pump's contamination accumulation has not yet significantly impacted current performance, but preventative cleaning is warranted.

[0070] Furthermore, based on the fusion results, cleaning decisions are made. The goal of cleaning decisions is to ensure the emulsion pump operates in optimal working condition. If a comprehensive assessment reveals that the pump's contamination status reaches a set threshold (for example, cumulative contamination exceeds a certain percentage, or the monitored status exceeds a certain contamination level), the system will initiate a cleaning process. Cleaning decisions do not rely solely on a single contamination indicator; instead, they combine historical and real-time data, taking into account factors such as the pump's operating efficiency, cleaning frequency, and operating costs, to intelligently determine whether cleaning is necessary.

[0071] Once a cleaning decision is made, the system initiates the corresponding cleaning control operation. The specific execution process of the cleaning control will vary depending on the contamination status of the pump and the cleaning requirements. The system can select different cleaning methods or cleaning intensities. For example, a lightly contaminated pump may only require a simple flush or adjustment, while a heavily contaminated pump may require more thorough cleaning, such as disassembly or deep cleaning. At this point, the system's intelligent control algorithm automatically adjusts the parameters of the cleaning process based on the cleaning strategy and the current contamination status to ensure that the cleaning effect achieves the desired goal. Through intelligent cleaning decision-making and control, the service life of the emulsion pump can be effectively extended, its operating efficiency can be optimized, and the waste of resources caused by excessive cleaning can be reduced.

[0072] Furthermore, the embodiment of the present application further includes step P80, which further includes:

[0073] P81: Record and store cleaning control records; P82: If the deviation between the accumulated pump state and the monitored pump state is greater than a preset value, feedback correction is performed on the k relative contamination rate mapping models based on the cleaning control records.

[0074] Optionally, the feedback mechanism of the intelligent cleaning control process can be further enhanced by recording cleaning control records and correcting the pollution rate mapping model based on state deviation to optimize the cleaning decision and control strategy of the pump and ensure the long-term effectiveness and accuracy of the cleaning operation.

[0075] Specifically, after each cleaning operation, detailed records of the cleaning process are kept and stored in cleaning control records. These records include the post-cleaning pump performance recovery, specific details of the cleaning operation (such as cleaning intensity and cleaning method), and the changes in the pump's status after cleaning. Post-cleaning performance recovery generally reflects the degree of functional recovery of the pump after the cleaning operation, for example, whether the pump's efficiency, energy consumption, and pressure have returned to near-ideal conditions. These cleaning control records not only provide data support for subsequent decision-making but also serve as an important basis for evaluating cleaning effectiveness.

[0076] By storing cleaning control records, the system can track operations over time and accumulate a wealth of cleaning data. These records provide valuable information for subsequent model optimization and contamination rate mapping corrections. For example, if pump performance fails to return to expected levels after cleaning, the system can analyze the possible causes and provide improvement strategies to ensure more accurate and effective cleaning operations in the future.

[0077] Furthermore, the model is fed back and corrected based on the deviation between the accumulated pump status and the monitored pump status. If the deviation exceeds a preset threshold, the system deems that the existing model's prediction deviates from the actual post-cleaning results, triggering a feedback correction mechanism. This deviation typically occurs when there's an error between the pump's actual operating conditions and the model's predictions. This can occur because certain pump characteristics aren't fully captured by the model, or because the post-cleaning recovery doesn't fully meet expectations.

[0078] Analysis of cleaning control records can identify potential model deficiencies. For example, if pump performance recovery after cleaning is incomplete, this could indicate that the contamination rate mapping model's predictions were inaccurate under certain operating conditions, resulting in suboptimal cleaning decisions. In this case, the system uses the recorded data to provide feedback and corrections to the k relative contamination rate mapping models, optimizing the model's parameters or structure to more accurately predict the pump's contamination rate and post-cleaning recovery. This dynamic optimization enables the system to continuously improve the management and maintenance of emulsion pumps, enhancing their operating efficiency and extending their service life.

[0079] Furthermore, the embodiment of the present application further includes step P90, which further includes:

[0080] P91: Based on the length of the operating condition interval, define k transition intervals of the standard operating condition interval; P92: Define continuity constraints, and according to the transition intervals and the continuity constraints, fuse the k relative pollution rate mapping models to obtain a relative pollution rate integrated mapping model; P93: According to the current cycle operation log and the relative pollution rate integrated mapping model, perform predictive analysis on the current cycle operation log to obtain the accumulated pump status.

[0081] In one possible embodiment of this application, by defining transition intervals and continuity constraints, k standard operating intervals are integrated to construct an integrated relative contamination rate mapping model. Predictive analysis is then performed based on the current cycle's operating logs to obtain the cumulative pump status. This process is designed to achieve smooth transitions between different operating intervals and improve the accuracy and robustness of the contamination rate mapping model, thereby providing a more accurate basis for subsequent cleaning decisions.

[0082] First, transition intervals are defined based on the length of each standard operating range. Each standard operating range represents the pump's state under different operating conditions. There may be a certain degree of transition between these ranges, indicating the continuity and complexity of the pump's state changes during transitions. To ensure smooth transitions between different operating ranges, the system defines transition intervals between these ranges.

[0083] For example, for cases of low complexity and a small number of segments, weighted fusion or continuity constraints can be used to achieve the connection of transition intervals. This method can ensure the smoothness and continuity of the transition interval by weighted averaging of multiple relative pollution rate mapping models. If a smooth connection with higher precision is required, mathematical methods such as basis function method or spline interpolation can be used. These methods can generate continuous and smooth transition curves to accurately connect different standard operating conditions. For example, spline interpolation can fit the curves between standard operating conditions intervals through polynomial functions, thereby achieving a more accurate transition. In some cases, if the target predicted by the model needs to be further optimized, the system can also consider adding regularization constraints. Regularization constraints help prevent model overfitting, ensure that the connection of the transition interval is not only smooth but also stable, and improve the generalization ability of the model in practical applications.

[0084] Furthermore, based on the defined transition intervals and continuity constraints, k relative pollution rate mapping models are fused to generate an integrated relative pollution rate mapping model. Continuity constraints ensure that the transitions between different standard operating intervals are not only smooth but also physically consistent. For example, the system can set constraints to ensure that the pollution rate trends between adjacent operating intervals remain consistent, without sudden or unreasonable jumps.

[0085] Through these constraints and the fusion process, an integrated pollution rate mapping model is generated that covers all operating ranges and achieves smooth transitions between them. This integrated model not only improves prediction accuracy but also adapts to changing operating environments, making pollution rate assessment more precise and stable.

[0086] After the integrated mapping model is completed, the cycle's operation log and the relative contamination rate integrated mapping model are used for predictive analysis. This cycle's operation log records the pump's specific operating data during the current cycle, including flow, pressure, temperature, and other information. By inputting this data into the integrated mapping model, the pump's contamination rate for the current cycle can be predicted and the pump's cumulative pump status can be further calculated.

[0087] The cumulative pump status is a cumulative calculation of the contamination level based on the predicted contamination rate for the current cycle. This status indicates the pump's contamination progress throughout the cycle and provides a basis for subsequent cleaning decisions. If the cumulative pump status reaches a preset threshold, the system triggers cleaning control to ensure the pump remains in optimal operating condition. This process enables the system to better adapt to changing operating environments, improving pump management efficiency and cleaning control accuracy.

[0088] In summary, the embodiments of the present application have at least the following technical effects:

[0089] This application calibrates the relative state curve of the emulsion pump, extracts data from the cleaning record database in combination with unique identification information, clusters and resamples the data, generates k working condition reorganization sample sets, and trains k pollution rate mapping models based on the reorganized sample data and the relative pump state curve to predict the cumulative pollution state of the pump. The pump's operating parameters are monitored in real time to evaluate the pollution state, and the cumulative pollution state and the monitored pump state are integrated to ultimately achieve accurate cleaning decision-making and control.

[0090] The technical effect of achieving precise cleaning control by dynamically adjusting cleaning operations through real-time monitoring and pollution status assessment, avoiding excessive or insufficient cleaning operations, and thus improving the operating efficiency of the pump and extending the life of the equipment has been achieved.

[0091] Example 2, based on the same inventive concept as the intelligent cleaning control method of an emulsion pump in the above embodiment, Figure 2 As shown, the present application provides an intelligent cleaning control system for an emulsion pump. The system and method embodiments in the present application are based on the same inventive concept. The system includes:

[0092] The relative pump state curve calibration module 11 is used to calibrate the relative pump state curve based on the target scene requirements and the target emulsion pump, wherein the relative pump state represents the pollution progress of the target emulsion pump.

[0093] The data extraction module 12 is used to obtain unique identification information of the target emulsion pump and extract data from the cleaning record database based on the unique identification information.

[0094] The clustering segmentation and resampling module 13 is used to perform clustering segmentation and resampling on the data extraction results to obtain reorganized sample data, wherein the data extraction results include sample pump control parameters and sample pump state parameters, and the reorganized sample data includes k working condition reorganized sample sets.

[0095] The rate mapping model construction module 14 is used to construct and train k relative pollution rate mapping models based on the recombined sample data and the relative pump state curve, wherein the relative pollution rate mapping model takes the pump control parameter as input.

[0096] The cumulative pump state prediction module 15 is used to extract the current cycle operation log of the target emulsion pump and predict the cumulative pump state of the target emulsion pump based on the current cycle operation log and the k relative pollution rate mapping models.

[0097] The direct contamination state evaluation module 16 is used to obtain the operating parameter information of the pump in real time, and perform direct contamination state evaluation based on the relative pump state curve to obtain the monitoring pump state.

[0098] The cleaning decision control module 17 is used to integrate the accumulated pump status and the monitoring pump status to perform cleaning decision and cleaning control.

[0099] Furthermore, the relative pump state curve calibration module 11 is further configured to perform the following steps:

[0100] Interactive target emulsion pump, obtain the key state indicator set of the pump; based on the target scenario requirements, determine the evaluation weight and performance control limit of the key state indicator set; define a relative performance evaluation function based on the evaluation weight and the key state indicator set, and calibrate the relative pump state curve of the target emulsion pump in combination with the performance control limit.

[0101] Furthermore, the cluster segmentation and resampling module 13 is further configured to perform the following steps:

[0102] The number of clusters k is determined based on the elbow method, and a density cluster analysis based on the control parameters of the sample pump is performed on the data extraction results; k standard operating condition intervals are defined according to the k clusters in the cluster analysis results; the data extraction results are segmented using the k standard operating condition intervals as segmentation constraints, and the segmentation results are output as k recombined data sets; based on preset confidence constraints, the k recombined data sets are confidence screened to obtain k recombined operating condition sample sets.

[0103] Furthermore, the rate mapping model construction module 14 is further configured to perform the following steps:

[0104] Based on the sample pump state parameters of the reorganized sample data; combining the relative pump state curve and the sample pump state parameters, the reorganized sample data is relatively state marked to obtain marked sample data; with the relative state marking result as supervision, k relative pollution rate mapping models are constructed and trained based on the marked sample data, wherein the k relative pollution rate mapping models correspond one-to-one to the k standard operating conditions.

[0105] Furthermore, the cumulative pump state prediction module 15 is further configured to perform the following steps:

[0106] Obtain the current cycle operation log of the target pump and extract the operating condition setting sequence from the current cycle operation log; perform mapping conversion on the operating condition setting sequence based on k relative pollution rate mapping models to obtain a relative pollution rate sequence; perform cumulative analysis on the relative pollution rate sequence to obtain the cumulative pump state.

[0107] Furthermore, the system further includes a feedback correction module, which is further configured to perform the following steps:

[0108] Record and store cleaning control records; if the deviation between the accumulated pump state and the monitored pump state is greater than a preset value, perform feedback correction on the k relative contamination rate mapping models based on the cleaning control records.

[0109] Furthermore, the system further includes an operation log prediction and analysis module, which is further configured to perform the following steps:

[0110] Based on the length of the operating condition interval, k transition intervals of the standard operating condition interval are defined; continuity constraints are defined, and according to the transition intervals and the continuity constraints, the k relative pollution rate mapping models are integrated to obtain a relative pollution rate integrated mapping model; according to the current cycle operation log and the relative pollution rate integrated mapping model, the current cycle operation log is predicted and analyzed to obtain the accumulated pump status.

[0111] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0112] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

[0113] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. An intelligent cleaning control method for an emulsion pump, characterized in that: The method comprises: Combine the target scenario requirements with the target emulsion pump to calibrate the relative pump status curve, where the relative pump status represents the contamination progress of the target emulsion pump, including: Interact with the target emulsion pump to obtain the key status indicator set of the pump; Based on the target scenario requirements, determine the evaluation weights and performance control limits of the key status indicator set; Defining a relative performance evaluation function according to the evaluation weight and the key status indicator set, and calibrating the relative pump status curve of the target emulsion pump in combination with the performance control limit; Obtaining unique identification information of a target emulsion pump, and extracting data from a cleaning record database based on the unique identification information; The data extraction results are clustered and resampled to obtain reorganized sample data, wherein the data extraction results include sample pump control parameters and sample pump state parameters, and the reorganized sample data includes k working condition reorganized sample sets, including: Determining the number of clusters k based on the elbow method, and performing density cluster analysis on the data extraction results based on the sample pump control parameters; Based on the k clusters in the cluster analysis results, k standard operating condition intervals are defined; Using the k standard operating condition intervals as segmentation constraints, segmenting the data extraction results, and outputting the segmentation results as k recombined data sets; Based on a preset confidence constraint, confidence screening is performed on the k reorganized data sets to obtain k working condition reorganized sample sets; Based on the recombined sample data and the relative pump state curve, k relative contamination rate mapping models are constructed and trained, wherein the relative contamination rate mapping model takes the pump control parameter as input, including: a sample pump state parameter based on the reorganized sample data; Combining the relative pump state curve with the sample pump state parameter, performing relative state marking on the recombined sample data to obtain marked sample data; Using the relative state labeling results as supervision, constructing and training k relative pollution rate mapping models based on the labeled sample data, wherein the k relative pollution rate mapping models correspond one-to-one to the k standard operating condition intervals; Extracting the current cycle operation log of the target emulsion pump, and predicting the cumulative pump state of the target emulsion pump based on the current cycle operation log and the k relative pollution rate mapping models, including: Obtaining the current cycle operation log of the target pump, and extracting the working condition setting sequence from the current cycle operation log; Performing mapping conversion on the operating condition setting sequence based on the k relative pollution rate mapping models to obtain a relative pollution rate sequence; Performing cumulative analysis on the relative pollution rate sequence to obtain the cumulative pump state; Acquire the operating parameter information of the pump in real time, and perform direct pollution status evaluation based on the relative pump status curve to obtain the monitoring pump status; The accumulated pump status and the monitored pump status are integrated to perform cleaning decision and cleaning control.

2. The intelligent cleaning control method for an emulsion pump according to claim 1, characterized in that: The method further comprises: Record and store cleaning control records; If the deviation between the accumulated pump state and the monitored pump state is greater than a preset value, feedback correction is performed on the k relative contamination rate mapping models based on the cleaning control record.

3. The intelligent cleaning control method for an emulsion pump according to claim 1, characterized in that: The method further comprises: Based on the length of the operating condition interval, defining k transition intervals of the standard operating condition interval; defining a continuity constraint, and fusing k relative pollution rate mapping models according to the transition interval and the continuity constraint to obtain a relative pollution rate integrated mapping model; According to the operation log of the current cycle and the relative pollution rate integrated mapping model, a prediction analysis is performed on the operation log of the current cycle to obtain the cumulative pump state.

4. The control system of the intelligent cleaning control method of the emulsion pump according to claim 1, characterized in that: The system comprises: A relative pump state curve calibration module is used to calibrate a relative pump state curve based on target scenario requirements and a target emulsion pump, wherein the relative pump state represents the pollution progress of the target emulsion pump, including: interacting with the target emulsion pump to obtain a key state indicator set of the pump; determining the evaluation weights and performance control limits of the key state indicator set based on the target scenario requirements; defining a relative performance evaluation function based on the evaluation weights and the key state indicator set, and calibrating the relative pump state curve of the target emulsion pump based on the performance control limits; A data extraction module, the data extraction module is used to obtain unique identification information of the target emulsion pump and extract data from the cleaning record database based on the unique identification information; A clustering segmentation and resampling module, the clustering segmentation and resampling module is used to perform clustering segmentation and resampling on the data extraction results to obtain reorganized sample data, wherein the data extraction results include sample pump control parameters and sample pump state parameters, and the reorganized sample data include k working condition reorganized sample sets, including: determining the number of clusters k based on the elbow method, and performing density clustering analysis on the data extraction results based on the sample pump control parameters; defining k standard working condition intervals based on the k clusters in the cluster analysis results; using the k standard working condition intervals as segmentation constraints, segmenting the data extraction results into segments, and outputting the segmented segmentation results as k reorganized data sets; performing confidence screening on the k reorganized data sets based on preset confidence constraints to obtain k working condition reorganized sample sets; A rate mapping model construction module, the rate mapping model construction module is used to construct and train k relative pollution rate mapping models based on the reorganized sample data and the relative pump state curve, wherein the relative pollution rate mapping model takes the pump control parameter as input, including: sample pump state parameters based on the reorganized sample data; relative state labeling of the reorganized sample data in combination with the relative pump state curve and the sample pump state parameters to obtain labeled sample data; using the relative state labeling result as supervision, constructing and training k relative pollution rate mapping models based on the labeled sample data, wherein the k relative pollution rate mapping models correspond one-to-one to the k standard operating condition intervals; A cumulative pump state prediction module is used to extract the current cycle operation log of the target emulsion pump and predict the cumulative pump state of the target emulsion pump based on the current cycle operation log and the k relative pollution rate mapping models, including: obtaining the current cycle operation log of the target pump and extracting a working condition setting sequence from the current cycle operation log; mapping and converting the working condition setting sequence based on the k relative pollution rate mapping models to obtain a relative pollution rate sequence; and performing cumulative analysis on the relative pollution rate sequence to obtain the cumulative pump state; A direct contamination state evaluation module, the direct contamination state evaluation module is used to obtain the operating parameter information of the pump in real time, and perform direct contamination state evaluation based on the relative pump state curve to obtain the monitoring pump state; A cleaning decision control module is used to integrate the accumulated pump status and the monitoring pump status to perform cleaning decision and cleaning control.

Images