Intelligent monitoring system and plastic centrifugal pump using same

Abstract

The invention relates to the technical field of mechanical equipment monitoring, and discloses an intelligent monitoring system which comprises a sensor module used for collecting vibration, temperature, pressure, flow and current data of a plastic centrifugal pump in real time; the embedded controller is connected with the sensor module and used for receiving and processing operation data in real time, and signal processing, preliminary data analysis and feature extraction are included; the data center module is connected with the embedded controller and is used for further extracting characteristic values of the operation data based on a deep learning algorithm, comparing the characteristic values with data in a historical database and generating fault early warning information; and the front-end interface module is connected with the data center module and is used for receiving and displaying the fault early warning information for operators to check. The sensor module is used for collecting vibration, temperature, pressure, flow and current data in real time, signal preprocessing is carried out through the embedded controller, feature values are extracted, and the adaptive capacity to complex working conditions of the pump body is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical equipment monitoring, in particular to an intelligent monitoring system and a plastic centrifugal pump using the system. Background Art

[0002] Currently, in industrial production, condition monitoring of mechanical equipment has become a crucial tool for ensuring safety and improving efficiency. Plastic centrifugal pumps, due to their excellent corrosion resistance and lightweight design, are widely used in fields such as nuclear power, pharmaceuticals, and new energy. As key equipment in liquid transportation, their stability and reliability directly impact the normal operation of the entire production line.

[0003] In the field of new energy, plastic centrifugal pumps, as key fluid conveying equipment, are widely used in scenarios such as energy storage system cooling, photovoltaic power station liquid circulation, hydrogen energy production and storage.

[0004] Due to the diverse physical and chemical properties of fluid media in new energy applications, such as coolant in energy storage systems, corrosive solutions in photovoltaic power plants, and high-pressure circulating water in the hydrogen energy sector, these special operating conditions place higher demands on pump vibration, pressure, and flow monitoring. However, existing technologies often rely on monitoring a single sensor or parameter, unable to simultaneously capture the complex coupling relationships between multiple parameters, resulting in incomplete predictions of potential failure risks.

[0005] Furthermore, single alarm mechanisms are prone to false or missed alarms when cooling demand fluctuates in energy storage systems or when the liquid circulation load in photovoltaic power plants changes. Furthermore, traditional monitoring solutions lack the ability to detect minor anomalies early on, resulting in failures often being discovered only at a critical stage. This can cause system downtime and impact the stable operation of the corresponding equipment. Summary of the Invention

[0006] In view of the shortcomings of the existing technology, the present invention provides an intelligent monitoring system and a plastic centrifugal pump using the system, which solves the problems of the existing technology in multi-dimensional monitoring, electrostatic safety, and early warning of faults in plastic centrifugal pumps.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: an intelligent monitoring system, comprising: Sensor module, used to collect vibration, temperature, pressure, flow and current data of plastic centrifugal pumps in real time; an embedded controller, connected to the sensor module, for receiving and processing operating data in real time, including signal processing, preliminary data analysis, and feature extraction; A data center module is connected to the embedded controller and is used to further extract characteristic values ​​of the operating data based on a deep learning algorithm, compare them with the data in the historical database, and generate fault warning information; A front-end interface module, connected to the data center module, for receiving and displaying fault warning information for operators to view; The alarm and storage module is connected to the data center module and the front-end interface module, and is used to trigger an alarm when the failure risk approaches a threshold and upload historical data to a remote storage platform.

[0008] Preferably, the sensor module includes: A vibration sensor is arranged near the bearing of the plastic centrifugal pump to monitor the vibration signal of the bearing; Temperature sensors are placed around the bearings and pump body of the plastic centrifugal pump to monitor the heat rise signals of the bearings; The pressure sensor is arranged in the inlet and outlet pipes of the plastic centrifugal pump to monitor the changes in liquid pressure; A flow sensor is arranged at the pump outlet to monitor the flow fluctuation of the liquid; The current sensor is arranged in the motor power supply circuit of the plastic centrifugal pump and is used to monitor the load changes of the motor.

[0009] Preferably, the embedded controller includes: A data receiving unit, used for receiving operation data sent by the sensor module; A signal processing unit, used for performing low-pass filtering and normalization processing on the operating data; The preliminary analysis unit is used to extract the initial characteristic values ​​of the vibration signal, including the root mean square value and signal energy.

[0010] Preferably, the data center module includes: A deep learning unit, used to further extract feature values ​​from operating data based on a specially designed deep learning algorithm; The data comparison unit is used to compare the characteristic value with the data in the historical database to generate fault warning information.

[0011] Preferably, the front-end interface module includes: An information receiving unit, configured to receive fault warning information generated by the data center module; The information display unit is used to display fault warning information to the operator in a visual manner.

[0012] Preferably, the alarm and storage module includes: An alarm unit, used to trigger an audible and visual alarm when the fault warning information indicates that the risk value is close to or exceeds a threshold; The storage unit is used to record operating data and fault warning information, and upload them to the remote cloud storage platform through the network interface.

[0013] An intelligent monitoring system is applied to a plastic centrifugal pump coated with an antistatic coating. The antistatic coating comprises a conductive polymer material or other material with antistatic properties, which is sprayed on the inner wall of the plastic centrifugal pump to reduce the accumulation of static electricity generated during liquid flow.

[0014] An intelligent monitoring method comprises the following steps: S1. Collect operating data of the plastic centrifugal pump through vibration sensors, temperature sensors, pressure sensors, flow sensors and current sensors; S2, sending the operating data to the embedded controller for signal preprocessing, including filtering, noise reduction and feature value extraction; S3. Upload the processed operation data to the data center and further extract feature values ​​based on the deep learning algorithm; S4. Compare the extracted characteristic values ​​with the data in the historical database to generate fault warning information; S5. Feedback the fault warning information to the front-end interface for operators to view; S6. When the fault risk approaches the set threshold, an alarm is triggered and the operating data is uploaded.

[0015] Preferably, the deep learning algorithm in step S3 includes the following processing steps: S31, further extracting characteristic values ​​from the operating data, including root mean square value, signal energy distribution and kurtosis parameter; S32. Match the extracted feature values ​​with abnormal patterns in the historical database to generate fault warning information.

[0016] Preferably, in the step S4, the data comparison is based on the difference calculation between the characteristic values ​​of the operating data and the health status data in the historical database, so as to generate fault warning information that conforms to the operating status.

[0017] The present invention provides an intelligent monitoring system and a plastic centrifugal pump using the system. It has the following beneficial effects: 1. The present invention uses a sensor module to collect vibration, temperature, pressure, flow and current data in real time, performs signal preprocessing and extracts characteristic values ​​through an embedded controller, and then combines the deep learning algorithm of the data center module to conduct a comprehensive analysis of the operating status, significantly improving the adaptability to complex working conditions of the pump body.

[0018] 2. This invention applies an antistatic coating to the inner wall of the plastic centrifugal pump, effectively reducing the risk of static electricity accumulation caused by friction during liquid flow. Compared with the existing method of dealing with static electricity only through external grounding, this invention solves the problem of inability to completely eliminate static electricity on the inner wall. This not only reduces the risk of spark discharge, but also prevents static electricity from interfering with pump operation and monitoring data, thereby improving the safety and reliability of plastic centrifugal pumps in scenarios where flammable liquids or highly corrosive media are transported.

[0019] 3. This invention provides early warning of potential failures in plastic centrifugal pumps through fault prediction and a graded alarm mechanism. Specifically, when the failure risk approaches a set threshold, an audible and visual alarm is triggered and the operator is remotely notified. Compared to the passive alarm methods used in existing technologies, this invention addresses the problem of delayed alarms leading to equipment damage or even shutdown, effectively reducing the unplanned downtime rate during plastic centrifugal pump operation and extending the equipment's service life.

[0020] 4. A modular storage solution is adopted. By storing real-time data, historical analysis results, and fault records in layers and supporting cloud synchronization, it enables long-term storage and traceability of equipment operation data. Compared with existing local single storage solutions, this invention effectively solves the problems of easy data loss and poor scalability, while also enhancing the system's analysis and optimization capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a schematic diagram of the monitoring system architecture of the present invention; Figure 2 This is a schematic diagram of the sensor module architecture of the present invention; Figure 3 is a schematic diagram of the embedded controller architecture of the present invention; Figure 4 This is a schematic diagram of the data center module architecture of the present invention; Figure 5 This is a schematic diagram of the front-end interface module architecture of the present invention; Figure 6 This is a schematic diagram of the alarm and storage module architecture of the present invention; Figure 7 Schematic diagram of the monitoring method steps of the present invention; Figure 8 Schematic diagram of the detailed steps of S3 of the present invention. DETAILED DESCRIPTION

[0022] The following will clearly and completely describe the technical solution of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0023] Example 1: Please see the attached Figure 1 , an embodiment of the present invention provides an intelligent monitoring system, comprising: Please see the attached Figure 2 ,sensor module, used to collect vibration, temperature, pressure, flow and current data of plastic centrifugal pump in real time; Specifically, in the intelligent monitoring system of the present invention, the sensor module is a core component for collecting real-time operating data from plastic centrifugal pumps. This module is directly connected to the embedded controller and provides basic data support for subsequent data processing, fault feature extraction, and early warning generation. The design of the sensor module directly determines the accuracy and stability of the overall monitoring system.

[0024] Typically, sensor modules collect real-time data on multiple operating parameters, including vibration, temperature, pressure, flow, and current, to provide high-precision data on the current operating status of the plastic centrifugal pump. This data is transmitted to an embedded controller for preliminary processing before being uploaded to a data center for in-depth analysis. Specifically, the placement of the sensor modules must be optimized based on the structural characteristics of the plastic centrifugal pump and the types of potential failures, taking into account the different parameter collection requirements.

[0025] In this embodiment, the sensor module mainly comprises the following categories: As an implementation method, a vibration sensor is installed near the bearing of the plastic centrifugal pump to monitor the vibration signal of the bearing in real time during operation.

[0026] In some embodiments, the vibration sensor uses a highly sensitive MEMS accelerometer with a sensitivity range that covers the vibration amplitudes that may occur in plastic centrifugal pumps. The high-frequency response characteristics of such sensors can accurately capture subtle changes in vibration signals.

[0027] Alternatively, temperature sensors are placed around the bearings and pump body to monitor real-time temperature changes. Abnormally high bearing temperatures can often be caused by lubrication failure, increased friction, or other issues. In one possible implementation, the temperature sensor uses a PT100 platinum resistance sensor, which has a temperature measurement range of -50°C to 250°C and offers high linearity and low drift. Temperature data is often combined with vibration data to generate a multi-dimensional fault signature.

[0028] Pressure sensors are placed on the inlet and outlet pipes of a plastic centrifugal pump to detect dynamic changes in liquid pressure. Specifically, the pressure at the pump inlet can indicate whether there is fluid blockage in the system, while fluctuations in outlet pressure may indicate an anomaly with the pump impeller or seals. As an implementation, the pressure sensor uses a silicon piezoresistive pressure sensor with a range of 0-10 MPa and high stability and corrosion resistance.

[0029] In some embodiments, a flow sensor is installed at the pump outlet to monitor electrolyte flow fluctuations. The pump's flow characteristics directly impact the system's output efficiency. As a possible alternative, an electromagnetic flowmeter can be used as the flow sensor, offering an accuracy of ±0.5% and suitable for flow measurement of conductive liquids such as electrolytes. Combining flow and pressure data can more effectively assess the pump's overall operating status.

[0030] Furthermore, a current sensor is placed in the motor power circuit of the plastic centrifugal pump to monitor motor load changes in real time. Specifically, current data is collected by a Hall-effect current sensor and combined with vibration and temperature data to form a comprehensive assessment of the pump's operating load. In one possible implementation, the current sensor has a measurement range of 0-100A and a response time of less than 1ms to accommodate the rapidly changing load characteristics of the motor.

[0031] In some extended implementations, the sensor module may further include: Humidity sensor, used to monitor the humidity changes in the pump chamber to determine whether there is leakage in the seal; The torque sensor is used to directly measure the torque of the motor shaft and further analyze the operating status of the pump.

[0032] Generally, the above sensors can be connected to the embedded controller via the CAN bus to achieve synchronous data collection and real-time transmission.

[0033] In short, the rational arrangement of sensor modules and high-precision data acquisition are the foundation of the intelligent system. Through the coordinated monitoring of multi-dimensional parameters, combined with formula calculation and feature extraction, the operating status of the plastic centrifugal pump can be fully reflected.

[0034] Please see the attached Figure 3 ,The embedded controller is connected to the sensor module and is used to receive and process the operation data in real time, including signal processing, preliminary data analysis and feature extraction; Specifically, the main responsibility of the embedded controller is to receive real-time data from the sensor module, perform preliminary signal processing and feature extraction, and pass the processing results to the data center module for further analysis.

[0035] Typically, an embedded controller connects to the sensor module via the CAN bus, receiving a variety of real-time operational data, including vibration, temperature, pressure, flow, and current signals. This data is first preprocessed within the controller, including filtering, noise reduction, and signal normalization. The embedded controller then performs feature extraction operations, such as calculating the root mean square value, signal energy, and other characteristic values, providing preliminary fault diagnosis capabilities.

[0036] In one possible implementation, the vibration signal received by the embedded controller usually contains the device operation signal and environmental noise. Therefore, a low-pass filter is set inside the embedded controller for noise reduction processing. The mathematical expression is: ; in: : The filtered signal, : The original signal collected by the sensor, : Impulse response function of the filter.

[0037] The filter's cutoff frequency is selected based on the vibration signal's frequency range to preserve key characteristic signals while eliminating high-frequency noise. Specifically, the embedded controller normalizes the filtered signal to eliminate the influence of amplitude differences between different signal sources on the analysis results. The embedded controller also extracts eigenvalues ​​from the normalized data. Common eigenvalues ​​for vibration signals include the root mean square value and signal energy.

[0038] The embedded controller is also responsible for extracting eigenvalues ​​from the normalized data. For vibration signals, commonly used eigenvalues ​​include the root mean square value and signal energy.

[0039] In some embodiments, the embedded controller also has a data caching function. When communication with the sensor module or data center module is interrupted, the embedded controller can temporarily store the received data and upload it after communication is restored. In terms of hardware design, the main components of the embedded controller include a high-performance microprocessor and a multi-channel analog-to-digital converter (ADC). The microprocessor needs to have a high main frequency (such as above 300MHz) to support the real-time operation of complex algorithms. The ADC module needs to have high precision (such as 16 bits) and a high sampling rate (such as above 10kHz) to ensure accurate collection of sensor data.

[0040] In some implementations, the embedded controller also supports multiple communication protocols, such as CAN, Ethernet, and RS485, to adapt to diverse industrial application scenarios.

[0041] Please see the attached Figure 4,The data center module is connected to the embedded controller and is used to further extract the characteristic values ​​of the operating data based on the deep learning algorithm, and compare it with the data in the historical database to generate fault warning information; Specifically, the main functions of the data center module include receiving preliminary processed data uploaded by the embedded controller, performing deep learning analysis, feature comparison and anomaly judgment on the data, and generating fault warning information for display on the front-end interface module.

[0042] Typically, the data center module connects to an embedded controller via a standard communication interface, receiving pre-processed data on vibration signals, temperature, pressure, flow, current, and other multi-dimensional parameters. This data is further analyzed within the data center module, including deep feature extraction, comparative analysis with historical databases, and anomaly detection and fault pattern recognition based on deep learning algorithms.

[0043] In one possible implementation, the data center module first formats the received multidimensional data and reorganizes the normalized data uploaded by the embedded controller into a data matrix suitable for deep learning algorithm input. The input matrix can be expressed as: ; Where: X: data input matrix; X i,j : the normalized value of the i-th feature in the j-th time window; m: the number of features; n: the number of time windows.

[0044] Specifically, the deep learning model in the data center module is trained offline to identify various failure modes of plastic centrifugal pumps. During training, faulty data and normal data are proportionally distributed as the training set, with labels representing different fault types.

[0045] The training goal is to minimize the loss function, which can be in the form of cross entropy loss: ; Where: L: loss function value; N: number of training samples; C: number of categories; y i,j : The true label of the i-th sample; : The predicted probability of the i-th sample.

[0046] Through loss function optimization, the model can achieve accurate classification of normal and fault states.

[0047] In some embodiments, the data center module further improves the accuracy of fault detection by comparing and analyzing the historical database. For each operating characteristic value, the difference between it and the historical health state mean is calculated, and the formula is: ; Where: D: difference; X i : current eigenvalue; : the mean of the corresponding feature in the historical database; m is the number of features. If the difference exceeds the preset threshold, the data center module will mark the data as abnormal and further analyze the cause of the abnormality.

[0048] In one possible implementation, the data center module also generates fault warnings. These warnings are generated based on the output of the deep learning model, combined with specific rules to produce easily understood alarm messages. For example, if the output probability of a particular fault mode exceeds 90%, the system can generate a specific warning message, such as "High risk of impeller failure, please check impeller operating status."

[0049] As an extension, the data center module can also upload analysis results to a cloud storage platform to facilitate the long-term preservation and retrieval of historical data. Cloud data can be used for further big data analysis and modeling, further optimizing fault diagnosis algorithms.

[0050] Please see the attached Figure 5 ,The front-end interface module is connected to the data center module and is used to receive and display fault warning information for operators to view; Specifically, the front-end interface module's primary function is to intuitively present fault warning information, operating status parameters, and historical trend data generated by the data center module to users. This module facilitates timely access to critical information and the ability to take appropriate action. The front-end interface module is closely integrated with the data center module, receiving warning results and analytical data from the data center via standard communication protocols.

[0051] Typically, the front-end interface module consists of a display device, a control unit, and interactive interface software. The display device displays the real-time status of multi-dimensional operating data such as vibration, temperature, pressure, flow, and current, and graphically displays fault warning information. The control unit manages data refresh and display logic.

[0052] In one possible implementation, the front-end interface module receives warning information and operational status data uploaded by the data center module via the TCP / IP protocol. The received data is processed by the parsing unit within the interface module and organized and classified according to a preset format. The basic process of data parsing can be expressed as follows: ; Where: D: parsed dataset; : Multidimensional parameters received from the data center module; f: Data parsing and classification function, including data format conversion and filtering operations.

[0053] As an option, the front-end interface module displays operating data and fault information in the form of visual charts. For example, the time domain changes of the vibration signal can be displayed using a line chart, while the root mean square (RMS) characteristics of different time windows can be intuitively compared using a bar chart. The coordinate relationship of the line chart is: ; Where: x: time point; y: vibration signal value at the corresponding time point.

[0054] Specifically, histograms are used to show the changing trend of specific characteristic values ​​(such as root mean square values), and their intuitiveness is more suitable for short-term trend analysis.

[0055] In some embodiments, the front-end interface module uses a tiered approach to display fault warning information. Generally, fault information is categorized into warning, fault, and emergency levels based on severity. Each level of information is displayed on the interface using a different color and symbol. For example, red indicates an emergency, yellow indicates a general warning, and green indicates normal operation.

[0056] As an extension, the front-end interface module supports playback and analysis of historical data. Users can select a specific time period through the interface to view the historical trends of various operating parameters.

[0057] In another possible implementation, the front-end interface module also has an exception report generation function. Based on the failure mode information provided by the data center module, this function outputs the exception data and analysis results in the form of a report. The report content includes the failure type, occurrence time, characteristic value change trend, and recommended measures.

[0058] Specifically, the interactive design of the front-end interface module also supports user customization of parameters such as alarm thresholds and display modes. For example, users can set the alarm threshold T for the fault prediction probability to adapt to the operating conditions of different equipment. The system's customization function is implemented through an interactive menu, and the parameter setting logic is as follows: ; in: : user-adjusted threshold; : default threshold; : User-defined adjustment amount.

[0059] Please see the attached Figure 6 ,The alarm and storage module is connected with the data center ,module and the front-end interface module.,It is used to trigger an alarm and upload historical data to the ,remote storage platform when the failure risk approaches the threshold.

[0060] Specifically, the main function of the alarm and storage module is to classify and process the fault warning information provided by the data center module, trigger the sound and light alarm signal, and store the operation data and analysis results.

[0061] Typically, the alarm and storage module receives fault warning information and related operating parameters from the data center module via standard communication protocols. Depending on the severity of the warning information, the alarm and storage module triggers different levels of alarm signals, such as low-risk warnings, high-risk alarms, or emergency failure notifications. The storage module is responsible for storing system operating data, fault signatures, and warning records in a structured format in a local or cloud database, and supports subsequent data export and analysis.

[0062] In one possible implementation, the alarm module triggers an alarm signal based on the failure prediction probability provided by the data center module. The alarm logic can be described as follows: ; Among them: A: alarm level; : predicted probability of failure mode, : Alarm threshold, corresponding to the dividing points of low-level and high-level alarms respectively.

[0063] Specifically, when the failure probability Exceeding the advanced threshold When the system is in emergency, it will trigger an emergency alarm, usually accompanied by sound and light signals.

[0064] Optionally, the alarm module can transmit alarm signals through a variety of output devices. For example, an audible alarm can be emitted through a buzzer or speaker, while a visual alarm can be indicated by a flashing LED or illuminated warning light. Typically, the frequency of the audible and visual alarm flashes is dynamically adjusted based on the alarm level. For example, for an emergency alarm, the audible frequency can be set to 1kHz, and the visual flashing frequency can be set to 2Hz.

[0065] In some embodiments, the alarm module also supports remote notification. Specifically, when the system triggers a high-risk or emergency alarm, the alarm module sends the warning information to a designated mobile device or cloud platform via the wireless communication module.

[0066] The storage module is responsible for recording system operation data and warning information to support subsequent historical trend analysis and anomaly tracing. Generally, the storage module adopts a hierarchical structure design, storing real-time data, fault records, and historical analysis results in different logical partitions. For example: ; Among them: S: storage data structure; R: real-time data; F: fault record; H: historical analysis results.

[0067] As an implementation method, the real-time data in the storage module is saved in the form of time series, and its structured representation is as follows: ; in: : Real-time data record at time t; : The multidimensional operating parameters corresponding to this time point.

[0068] Fault records include information such as fault type, occurrence time, characteristic value change trend, and warning level. Its storage format can be further expanded to support big data analysis or artificial intelligence model training.

[0069] As an extension, the storage module also supports local and cloud data synchronization. Generally, real-time data is saved to the local storage device first, while key fault information and historical analysis results are uploaded to the cloud platform via the wireless communication module. Cloud storage not only supports multi-device access but also enables further optimization of fault diagnosis strategies through big data analysis.

[0070] The above system is applied to a plastic centrifugal pump coated with an antistatic coating. The antistatic coating includes a conductive polymer material or other material with antistatic properties, which is sprayed on the inner wall of the plastic centrifugal pump to reduce the accumulation of static electricity generated during the flow of liquid.

[0071] Specifically, in general, antistatic coatings are made of conductive polymer materials or other materials with antistatic properties. These materials can significantly reduce the surface resistance of the inner wall of the plastic, so that static electricity can be quickly introduced into the grounding system to avoid accumulation on the inner wall. As an implementation method, the main materials of antistatic coatings include conductive carbon nanotubes, graphene or polymers doped with conductive fillers. For example, epoxy resin coatings doped with carbon nanotubes not only have excellent conductive properties, but also maintain high chemical stability, which is suitable for working conditions of corrosive media in plastic centrifugal pumps. The surface resistance range of the coating is usually controlled in the range of , which can effectively remove static electricity without affecting the insulation performance of the pump body. This range is adjusted according to the application scenario of the pump and the conductivity of the conveyed liquid.

[0072] In one possible implementation, the antistatic coating is evenly coated on the inner wall of the plastic centrifugal pump through a spraying process. The spraying process needs to meet the following key requirements: The coating thickness is generally controlled in the range of 10~50μm to ensure sufficient conductivity and durability.

[0073] The adhesion of the coating must meet the ASTM D3359 standard (usually requiring adhesion to reach 4B or above) to prevent the coating from peeling off under the erosion of high-flow liquids.

[0074] After spraying, the coating needs to undergo a heat treatment or curing process, such as heating to 80°C~120°C for heat curing, to enhance the bonding strength between the coating and the substrate.

[0075] During the spraying process, ensure that the coating material evenly covers all surfaces of the inner wall of the pump cavity, especially the curved surface of the pump body, near the impeller and other areas prone to static electricity accumulation.

[0076] Specifically, antistatic coatings can significantly reduce the following risks by reducing static electricity buildup: Spark discharge risk: When conveying flammable liquids (such as organic solvents), static electricity buildup can cause spark discharges, resulting in explosions or fires. Antistatic coatings can quickly dissipate static electricity, minimizing this risk.

[0077] Signal interference: Static electricity can interfere with data collection from vibration or current sensors, affecting the system's monitoring accuracy. Antistatic coatings can reduce the impact of this type of electromagnetic interference on monitoring systems.

[0078] Equipment damage: Static electricity accumulation may cause surface aging, cracking, or local damage to plastic materials. Coating protection can extend the service life of the equipment.

[0079] Example 2: Please see the attached Figure 7-Figure 8 Based on the first embodiment, this embodiment provides an intelligent monitoring method, including the following steps: S1. Collect operating data of the plastic centrifugal pump through vibration sensors, temperature sensors, pressure sensors, flow sensors and current sensors; Specifically, during the operation of a plastic centrifugal pump, parameters such as vibration, temperature, pressure, flow, and current are important indicators of the pump's operating status. Vibration sensors are placed at key locations on the bearings and pump body to monitor potential mechanical vibration changes during operation, such as abnormal vibration caused by bearing wear or impeller imbalance. Temperature sensors are primarily located near the bearings and on the outer wall of the pump body to monitor temperature rise in the bearings and within the pump chamber, identifying any risk of lubrication deficiency or other overheating.

[0080] Pressure sensors installed in the pump's inlet and outlet pipes monitor dynamic changes in fluid pressure in real time to identify issues such as pipe blockage, pump leakage, or abnormal impeller wear. Flow sensors installed at the pump outlet capture flow fluctuations, particularly significant drops in flow, to indicate possible blockage or damage. Current sensors integrated into the motor's power supply circuit monitor real-time changes in motor load, helping to determine whether operating power is abnormal.

[0081] Typically, these sensors collect data at fixed time intervals (such as milliseconds or seconds) and output the collected raw signals to an embedded controller.

[0082] S2, sending the operating data to the embedded controller for signal preprocessing, including filtering, noise reduction and feature value extraction; Specifically, collected operating data often contains various noise components, such as environmental noise and electromagnetic interference. After receiving the data, the embedded controller first uses filtering technology to reduce noise on the raw signal, preserving the core signal characteristics while eliminating high-frequency or random noise. This filtered data more accurately reflects the actual operating status of the pump.

[0083] After signal noise reduction is complete, the embedded controller performs feature extraction on the data. The goal of feature extraction is to extract key quantifiable indicators from complex raw data, such as the root mean square (RMS) value of the vibration signal, peak temperature, pressure fluctuation amplitude, and instantaneous current fluctuation. These feature values ​​serve as core data for determining the health of the pump.

[0084] S3. Upload the processed operation data to the data center and further extract feature values ​​based on the deep learning algorithm; S31, further extracting characteristic values ​​from the operating data, including root mean square value, signal energy distribution and kurtosis parameter; S32. Match the extracted feature values ​​with abnormal patterns in the historical database to generate fault warning information.

[0085] Specifically, the eigenvalues ​​collected by the embedded controller provide only basic analysis, and the data center module further processes these eigenvalues ​​through deep learning. After data is uploaded, the data center uses a pre-trained deep learning model to extract deeper feature relationships from the eigenvalues, such as nonlinear correlations between operating modes or dynamic changes between multi-dimensional features.

[0086] For example, when vibration, temperature, and pressure characteristics fluctuate abnormally simultaneously, deep learning algorithms can identify the likely fault type based on the coupling relationship between these characteristics. The data center module can also detect patterns that are difficult for traditional algorithms to capture, such as frequent but subtle vibration anomalies that may indicate an early stage of bearing failure.

[0087] The core goal of this process is to extract high-level features that are meaningful for fault diagnosis from complex data and enhance the accuracy of subsequent analysis.

[0088] S4. Compare the extracted characteristic values ​​with the data in the historical database to generate fault warning information; Specifically, after completing the deep learning analysis, the data center module compares the extracted feature values ​​with the data in the historical database. The historical database stores the feature value distribution under normal operating conditions and the feature templates of various typical failure modes.

[0089] By comparing the values, the system can quickly determine whether the current characteristic value deviates from the normal range and whether the degree of deviation is sufficient to constitute a potential risk. For example, if the frequency component of the vibration signal closely matches the characteristic frequency of a bearing fault, the system will determine that there is a risk of bearing failure.

[0090] Fault warning information includes, but is not limited to, the potential fault type (e.g., impeller anomaly, bearing wear), risk level (low, medium, high), and recommended action. This information is then packaged into a visual data package for display on the front-end interface module.

[0091] S5. Feedback the fault warning information to the front-end interface for operators to view; Specifically, after receiving fault warning information, the front-end interface module displays it visually to the operator. Specifically, the interface module uses different colors, graphics, or text to indicate the fault type and risk level. For example, red indicates an emergency fault, yellow indicates a medium risk, and green indicates normal status.

[0092] The front-end interface also supports dynamic display of multi-dimensional data. For example, real-time curves of vibration, temperature, pressure, flow, and current can help operators intuitively understand equipment operating conditions. Operators can also use the interface to further query historical data or view detailed analysis results of abnormal data.

[0093] The goal of this process is to help operators quickly understand the status of the equipment and take emergency measures when necessary.

[0094] S6. When the fault risk approaches the set threshold, an alarm is triggered and the operating data is uploaded.

[0095] Specifically, when the risk level in the fault warning information reaches or exceeds the set alarm threshold, the alarm and storage module immediately triggers an alarm signal. Alarm signals typically include audible and visual alarms and remote notifications, ensuring that operators receive warnings as quickly as possible.

[0096] At the same time, the alarm and storage module records current operating data and analysis results to local storage devices and uploads them to cloud storage platforms. Cloud storage not only supports the long-term preservation of historical data but also provides a data foundation for subsequent big data analysis and model optimization.

[0097] The stored data includes real-time operating data, feature value extraction results, failure mode analysis, and system-generated alarm records. This data can be used to track the causes of failures, evaluate equipment operating trends, and provide a basis for subsequent maintenance decisions.

[0098] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent monitoring system, characterized in that: include: Sensor module, used to collect vibration, temperature, pressure, flow and current data of plastic centrifugal pumps in real time; an embedded controller, connected to the sensor module, for receiving and processing operating data in real time, including signal processing, preliminary data analysis, and feature extraction; A data center module is connected to the embedded controller and is used to further extract characteristic values ​​of the operating data based on a deep learning algorithm, compare them with the data in the historical database, and generate fault warning information; A front-end interface module, connected to the data center module, for receiving and displaying fault warning information for operators to view; The alarm and storage module is connected to the data center module and the front-end interface module, and is used to trigger an alarm when the failure risk approaches a threshold and upload historical data to a remote storage platform.

2. An intelligent monitoring system according to claim 1, characterized in that , the sensor module includes: A vibration sensor is arranged near the bearing of the plastic centrifugal pump to monitor the vibration signal of the bearing; Temperature sensors are placed around the bearings and pump body of the plastic centrifugal pump to monitor the heat rise signals of the bearings; The pressure sensor is arranged in the inlet and outlet pipes of the plastic centrifugal pump to monitor the changes in liquid pressure; A flow sensor is arranged at the pump outlet to monitor the flow fluctuation of the liquid; The current sensor is arranged in the motor power supply circuit of the plastic centrifugal pump and is used to monitor the load changes of the motor.

3. The intelligent monitoring system according to claim 1, characterized in that , the embedded controller includes: A data receiving unit, used for receiving operation data sent by the sensor module; A signal processing unit, used for performing low-pass filtering and normalization processing on the operating data; The preliminary analysis unit is used to extract the initial characteristic values ​​of the vibration signal, including the root mean square value and signal energy.

4. An intelligent monitoring system according to claim 1, characterized in that , the data center module includes: A deep learning unit, used to further extract feature values ​​from operating data based on a specially designed deep learning algorithm; The data comparison unit is used to compare the characteristic value with the data in the historical database to generate fault warning information.

5. The intelligent monitoring system according to claim 1, characterized in that , the front-end interface module includes: An information receiving unit, configured to receive fault warning information generated by the data center module; The information display unit is used to display fault warning information to the operator in a visual manner.

6. An intelligent monitoring system according to claim 1, characterized in that ,The alarm and storage module includes: An alarm unit, used to trigger an audible and visual alarm when the fault warning information indicates that the risk value is close to or exceeds a threshold; The storage unit is used to record operating data and fault warning information, and upload them to the remote cloud storage platform through the network interface.

7. The intelligent monitoring system according to claim 1 is applied to a plastic centrifugal pump coated with an antistatic coating, characterized in that ,The antistatic coating includes a conductive polymer material or other ,material with antistatic properties, which is sprayed on the inner wall of the plastic centrifugal ,pump to reduce the accumulation of static electricity generated during the ,flow of liquid.

8. An intelligent monitoring method, according to an intelligent monitoring system according to any one of claims 1 to 7, characterized in that , including the following steps: S1. Collect operating data of the plastic centrifugal pump through vibration sensors, temperature sensors, pressure sensors, flow sensors and current sensors; S2, sending the operating data to the embedded controller for signal preprocessing, including filtering, noise reduction and feature value extraction; S3. Upload the processed operation data to the data center and further extract feature values ​​based on the deep learning algorithm; S4. Compare the extracted characteristic values ​​with the data in the historical database to generate fault warning information; S5. Feedback the fault warning information to the front-end interface for operators to view; S6. When the fault risk approaches the set threshold, an alarm is triggered and the operating data is uploaded.

9. An intelligent monitoring method according to claim 8, characterized in that ,The deep learning algorithm in step S3 includes the following processing steps: S31, further extracting characteristic values ​​from the operating data, including root mean square value, signal energy distribution and kurtosis parameter; S32. Match the extracted feature values ​​with abnormal patterns in the historical database to generate fault warning information.

10. An intelligent monitoring method according to claim 8, characterized in that ,In the S4 step, data comparison is based on the ,difference calculation between the characteristic values ​​of the operating ,data and the health status data in the historical database, and ,is used to generate fault warning information that is consistent with the ,operating status.

Images