Sensor device for monitoring a centrifugal pump

Abstract

The present invention relates to a sensor device for monitoring a centrifugal pump, comprising at least one signal input unit including a current transformer input, an analog input, and / or a digital input, a low-power wide-area network (LPWAN) wireless interface, and a processor-based data processing device. The data processing device is configured to detect at least one monitoring signal characterizing the state of the centrifugal pump through the at least one signal input unit at least once per second, and is configured to perform analysis and data reduction preprocessing on the at least one monitoring signal in relation to the alarm state of the centrifugal pump, statistical evaluation of the state of the centrifugal pump, and / or monitoring and prediction of the state of the centrifugal pump. Additionally, the at least one preprocessed monitoring signal is transmitted via the LPWAN wireless interface at a frequency not exceeding once every five minutes to a cloud infrastructure configured for post-processing of the transmitted preprocessed monitoring signal.

Description

Technology Field

[0001] The present invention relates to a sensor device for monitoring a centrifugal pump, comprising at least one signal input unit, a Low Power Wide Area Network (LPWAN) wireless interface, and a processor-based data processing unit. Additionally, the present invention relates to a method for monitoring a centrifugal pump using the sensor device, comprising the step of detecting at least one monitoring signal characterizing the state of the centrifugal pump at at least one signal input unit of the sensor device.

[0003] The innovative technology of the present invention contributes to United Nations Sustainable Development Goal (SDG) 9 in particular by improving the efficiency and lifespan of infrastructure, contributes to SDG 11 in ensuring a stable water supply in urban and rural areas, and contributes to SDG 13 in reducing energy consumption and greenhouse gas emissions through optimized operation of pumps and energy-efficient data transmission. Background Technology

[0005] Centrifugal pumps are widely used in various industrial and commercial applications, including water and wastewater treatment, the chemical industry, and heating and cooling circuits. Since the functionality and efficiency of centrifugal pumps depend significantly on their operating conditions, continuous or periodic monitoring of operating parameters is required to detect failures or malfunctions early.

[0007] Conventional centrifugal pump monitoring systems are primarily based on wired sensor solutions integrated into a central control unit. While these monitoring systems enable detailed recording of operating parameters, they require extensive wiring, continuous power supply, and complex data processing. Furthermore, these systems often generate large volumes of data, and the transmission and analysis of this data require significant resources and storage space.

[0009] Recently, wireless monitoring solutions with improved installation flexibility and scalability are being developed. In particular, mobile communication-based technologies enable long-distance data transmission. However, these technologies face specific challenges in that they must transmit data with minimal energy consumption. One problem with conventional technology is that the high transmission frequency of measurement values ​​leads to unnecessary energy consumption and network load. The problem to be solved

[0011] Against the background described above, the objective of the present invention is to provide a sensor device for monitoring a centrifugal pump and a corresponding method that enables flexible installation, subsequent application to existing facilities, and energy-efficient wireless data transmission.

[0013] The objective of the present invention is achieved by the features described in the independent claim, and preferred embodiments are described in the dependent claim.

[0015] Accordingly, the objective of the present invention is achieved by a sensor device for monitoring a centrifugal pump, said sensor device comprising at least one signal input unit including a current transformer input unit, an analog input unit and / or a digital input unit, a low-power wide area network (LPWAN) wireless interface, and a processor-based data processing unit.

[0016] The above data processing device is,

[0017] Detecting at least one monitoring signal characterizing the state of the centrifugal pump at least once per second through the above at least one signal input unit, and

[0018] In particular, regarding the alarm status of the centrifugal pump, statistical evaluation of the centrifugal pump status, and / or monitoring and prediction of the centrifugal pump status, preprocessing for analysis and / or data reduction is performed on the at least one monitoring signal, and

[0019] The cloud infrastructure configured for post-processing of the pre-processed monitoring signal is configured to transmit at least one monitoring signal through the LPWAN wireless interface at a frequency not exceeding once every 5 minutes.

[0021] That is, the present invention relates to a sensor device for monitoring a centrifugal pump having a plurality of signal input units, including a current transformer input unit, an analog input unit, or a digital input unit. The sensor device includes an LPWAN wireless interface that enables energy-efficient wireless data transmission. A processor-based data processing unit of the sensor device is configured to receive and process at least one, preferably a plurality of, monitoring signals from the at least one signal input unit at short intervals of at least once per second. The recorded monitoring signals undergo preprocessing, particularly for data compression, and such preprocessing includes analysis of possible alarm states of the centrifugal pump, statistical evaluation of the pump state, and / or continuous state monitoring and prediction. After preprocessing, the monitoring signals are transmitted to a cloud infrastructure via the LPWAN wireless interface at defined intervals of at least 5 minutes. The cloud infrastructure is configured to postprocess the transmitted preprocessed monitoring signals, for example, to perform additional analysis or to initiate additional measures if necessary.

[0023] The sensor device for monitoring the centrifugal pump described above includes at least one signal input section that may be composed of a current transformer input section, an analog input section, and / or a digital input section, so as to be able to detect various types of monitoring signals, such as current values ​​for power monitoring, analog vibration signals for detecting mechanical abnormalities, or digital signals from the centrifugal pump. The current transformer input section is preferably provided to connect a current transformer (also referred to as a current sensor or current measuring transformer), thereby enabling the measurement of electrical current and the preparation of a signal for subsequent processing. The current transformer converts, for example, a high current flowing through the power connection cable of the centrifugal pump into a smaller proportional current or voltage that can be reliably and accurately detected at the signal input section of the sensor device.

[0025] One example of a monitoring signal characterizing the condition of a centrifugal pump is the monitoring of its motor current. If the motor of a centrifugal pump consumes more current than during normal operation, it may indicate mechanical problems such as shaft jamming, wear, or bearing defects. To reliably measure this motor current, a current transformer is mounted on one of the current-carrying conductors of the power connection cable. The current transformer generates an attenuated signal, which is input to a sensor device via the transformer's input for analysis. Another application is load monitoring in industrial pump systems, where power consumption serves as an indicator of the overall efficiency of the pump system. By continuously recording and analyzing power consumption, the sensor device can detect abnormal conditions early, improve the energy efficiency of the pump system, and contribute to reducing maintenance costs. An advantage of the current transformer input or the current transformer itself is that it enables non-contact measurement.

[0027] The LPWAN wireless interface enables energy-efficient wireless data transmission via protocols such as LoRaWAN, LTE-M, or NB-IoT, making the sensor device particularly suitable for retrofitting in remote or hard-to-access locations. The at least one monitoring signal describes the state of the centrifugal pump and may be, for example, current consumption to identify load changes or a frequency spectrum to detect cavitation phenomena. The at least one monitoring signal undergoes preprocessing for analysis and / or data reduction, and such preprocessing may include, for example, fault signal detection by pattern matching, statistical long-term evaluation of operating trends, or predictive maintenance prediction based on the detection of deviations from normal operating values.

[0029] In this specification, “analyzing pre-processing” means, in particular, that a sensor device examines an input monitoring signal to identify meaningful features. For example, the sensor device may detect power consumption patterns, vibration patterns, or temperature change patterns to identify signs indicating impending failure, such as imbalance, cavitation, or bearing jamming. Such analysis may be based on various methods, including frequency analysis, particularly Fast Fourier Transform (FFT), statistical methods such as mean values ​​or standard deviations, or threshold inspection methods such as the detection of a threshold increase in current.

[0031] In this specification, “data-reducing pre-processing” means, in particular, compressing recorded and analyzed monitoring signals to the extent that they can be processed and transmitted with minimal memory and energy consumption without losing relevant information. For example, instead of a continuous current signal with a high sampling rate, the sensor device may transmit only statistical parameters such as average current per minute, modified waveforms with reduced frequency resolution, or binary states such as “normal operation” or “error.” As an example, when monitoring the motor of a centrifugal pump, instead of transmitting the entire current curve to the cloud in real time, only whether current consumption is within a normal range or whether a significant increase has been detected may be reported.

[0033] Such analysis and / or data reduction preprocessing minimizes network load, extends the battery life of wireless sensors detecting at least one monitoring signal, and improves the efficiency of subsequent data analysis performed in the cloud. The preprocessed monitoring signal is transmitted to a cloud infrastructure via an LPWAN interface at a frequency not exceeding once every five minutes, and said cloud infrastructure is configured specifically for further analysis, long-term storage, and possible control measures. A processor-based data processing unit, specifically acting as a central computing module of the sensor device, is responsible for acquiring the monitoring signal, performing data reduction analysis to reduce the amount of data to be subsequently transmitted, and transmitting the monitoring signal in the form of reduced data. The processor-based data processing unit generally includes a microcontroller (MCU) or a microprocessor and is specialized to process the monitoring signal in real-time, filter only relevant information through preprocessing, and transmit only essential information.

[0035] Cloud infrastructure refers specifically to a networked, scalable IT platform accessible via the Internet or a private network, providing computing resources, storage space, and / or software services. In this sensor device, the cloud infrastructure functions as a central data processing and analysis platform that further processes received monitoring signals, performs long-term storage, and enables intelligent diagnostics. The cloud infrastructure enables high-performance data analysis algorithms for trend evaluation, historical comparison, and / or detection of critical anomalies, such as the imminent bearing failure caused by increased irregularities in power consumption. Post-processing may include statistical long-term evaluation of operating trends or predictive maintenance forecasts, or, in some cases, may be performed prior to pre-processing. When such anomalies are detected, the cloud can transmit automated maintenance signals to maintenance personnel or generate preemptive maintenance plans.

[0037] Furthermore, the cloud infrastructure can identify system-wide patterns or failure patterns across multiple locations by simultaneously storing and comparing pre-processed monitoring signals transmitted from hundreds of sensor devices. The cloud infrastructure may consist of a public cloud, a private cloud, or a hybrid cloud system. Advantages of the cloud infrastructure include high computational performance, scalability, automatic software updates, location-agnostic accessibility, and / or the ability to develop more accurate failure diagnosis and maintenance strategies for centrifugal pumps through artificial intelligence (AI) and machine learning. At the same time, local hardware requirements for sensor devices are reduced by migrating complex data processing and long-term storage to the cloud.

[0039] According to a preferred additional embodiment, the sensor device comprises two or three current transformer inputs, at least one analog output, at least one digital output, and / or an RS485 interface. The analog input is preferably configured for an input current of 4 to 20 mA and / or an input voltage of 0 to 10 V. The RS485 interface is an industry standard for a physical interface for asynchronous serial data transmission, also known as EIA-485 or RS-485. The RS485 interface is preferably configured according to the TIA-485-A standard, in particular ANSI / TIA / EIA-485-A-98 (R2003). Additionally, the sensor device preferably comprises a USB-C interface, particularly for programming, and the USB-C interface is preferably configured according to IEC 62680-1-3 (July 2016). These additional embodiments enable the sensor device to be multipurposely integrated into various applications and / or monitoring and control systems.

[0041] According to another preferred embodiment, the analysis and / or data reduction preprocessing comprises data compression and / or selection for at least one monitoring signal. Data compression may be performed, for example, by a lossless method such as delta coding that stores only the changes between consecutive monitoring signals, or by run-length coding that compresses sections where the same monitoring signal is repeated for a long time. Alternatively, a lossy compression method may be used that converts the time-based monitoring signal into a frequency spectrum using the Fast Fourier Transform (FFT) and stores and transmits only the relevant frequency components thereof.

[0043] The selection of monitoring signals can be performed by, for example, a threshold method that further processes only signals exceeding or falling below defined limits, or by an event-based selection method that transmits monitoring signals only when abnormal behavior, such as sudden vibration or power fluctuations in the centrifugal pump, is detected. Additionally, an adaptive selection method can be used in which the recording frequency varies depending on the operating state of the centrifugal pump; for example, a high recording frequency can be used when there is a significant change, and a low recording frequency can be used during normal operation. These methods improve the energy efficiency of the sensor device, reduce storage space, and optimize wireless communication, particularly by improving transmission efficiency through the LPWAN wireless interface.

[0045] According to another preferred embodiment, the analysis and / or data reduction preprocessing comprises amplitude reduction and / or time reduction of at least one monitoring signal. Amplitude reduction can be achieved by quantization, which maps a continuous monitoring signal to a limited number of levels to ignore minute changes, or by a smoothing method, such as a moving average filter, to remove small fluctuations and retain only the relevant monitoring signal. Alternatively, an amplitude clipping method that stores only signals exceeding or falling below a specific threshold can be used to detect only significant changes in the pump state.

[0047] Time reduction can be achieved, for example, by downsampling, which records monitoring signals only at longer time intervals, thereby reducing the amount of data generated and processed by increasing the recording interval, for example, from 1 ms to 100 ms. Another method is event-driven sampling, which records or processes monitoring signals only when specific changes in the signal curve, for example, pressure of a centrifugal pump, or sudden deviations in vibration, are detected. These embodiments significantly reduce the data load, extend the battery life of battery-powered sensor devices, and improve processing and transmission efficiency by optimizing the utilization of available bandwidth for wireless transmission via the LPWAN wireless interface.

[0049] According to another preferred embodiment, the data transmission rate of at least one monitoring signal is 300, 500, 1,000, or 2,000 times greater than the data transmission rate of at least one preprocessed monitoring signal being transmitted. This configuration enables significant data reduction. The sensor device can collect, analyze, and compress a large amount of monitoring signals in the form of raw data, and then transmit only the extremely reduced amount of data as a preprocessed monitoring signal to a cloud infrastructure. As an example, when monitoring the power consumption of a centrifugal pump, the monitoring signal may be recorded as a raw signal at a sampling rate of 10 kHz, but after preprocessing, only compressed state values ​​or statistical parameters, such as average values ​​or peak values ​​per minute, are transmitted as a preprocessed monitoring signal.

[0051] As another example, a vibration sensor can perform continuous measurements at 1 kHz, but summarized information can be transmitted as a preprocessed monitoring signal only when a threshold is reached or an anomaly occurs. Another example is the Fourier analysis of pump noise, where only significant frequency changes among millions of raw data points are extracted as monitoring signals and transmitted as preprocessed monitoring signals. This configuration improves wireless transmission efficiency over LPWAN while retaining only the information relevant to centrifugal pump monitoring, and reduces bandwidth and energy consumption through data reduction in the pre-transmission stage.

[0053] According to another preferred embodiment, at least one preprocessed monitoring signal transmitted is of a Boolean data type. After such additional preprocessing, the at least one monitoring signal may have only one of two states, for example, 0 or 1, true or false, or on or off. Since this configuration transmits only binary values ​​instead of complex raw data, it maximizes the data reduction effect and significantly reduces transmission time and energy consumption. As a specific example, in the case of detecting the alarm status of a centrifugal pump, the sensor device may continuously measure vibration, current consumption, or temperature and process these monitoring signals. Instead of transmitting high-resolution monitoring signals, a Boolean value indicating only whether a threshold value has been exceeded, for example, 1 is an alarm and 0 is a non-alarm, may be transmitted. As another example, in condition monitoring based on pattern evaluation, an internal model of the sensor device can analyze whether the centrifugal pump is in normal operation mode or failure mode and transmit only the result as a Boolean signal; for instance, “normal pump operation” can be represented as 0 and “maintenance required” as 1. This configuration is particularly suitable for energy-efficient LPWAN networks because it reliably transmits information necessary for decision-making while minimizing data load. In other words, this binary representation of the monitoring signal significantly reduces the complexity of the transmitted data.

[0055] According to another preferred embodiment, the LPWAN wireless interface is configured to transmit at least one preprocessed monitoring signal using a Long Range Wide Area Network (LoRaWAN) network protocol, a Long Term Evolution for Machines (LTE-M) network protocol, and / or a NarrowBand IoT (NB-IoT) network protocol. LoRaWAN is a wireless protocol specifically developed for the Internet of Things (IoT) that operates in an unlicensed frequency band and is optimized for long-distance communication with low energy consumption, enabling sensors to transmit monitoring signals over several kilometers to gateways, cloud infrastructure, etc., for example, to monitor a centrifugal pump in a remote industrial facility without wiring infrastructure.

[0057] LTE-M is based on the LTE mobile communication standard and offers higher bandwidth and lower latency than LoRaWAN, making it suitable for applications requiring near real-time monitoring of centrifugal pump vibrations in municipal waterworks. Meanwhile, Narrowband Internet of Things (NB-IoT) is a narrowband mobile communication technology with excellent penetration capabilities, particularly into building interiors or underground spaces, enabling reliable monitoring of centrifugal pumps installed in basements or shafts. Furthermore, NB-IoT offers high scalability, allowing multiple data processing facilities to be networked within a single network. By supporting these different LPWAN protocols, sensor devices can be flexibly applied to various use cases by selecting the appropriate transmission technology based on installation infrastructure and requirements.

[0059] According to another preferred embodiment, transmission is performed continuously and / or periodically, and / or in cycles not shorter than 10, 15, 20, 30, or 60 minutes. This additional configuration enables flexible adaptation to various monitoring requirements. Continuous transmission specifically means transmitting pre-processed monitoring signals in real-time or with minimal delay, which is particularly advantageous for safety-critical applications where an immediate alarm is required, for example, when power consumption of a centrifugal pump increases rapidly, indicating that a blockage is imminent.

[0061] On the other hand, periodic transmission is performed at fixed intervals, such as 30 minutes or 1 hour, and is suitable for applications that do not require an immediate response, such as long-term analysis of wear patterns or efficiency trends of centrifugal pumps. Additionally, by limiting the transmission frequency to minimum intervals such as 10 minutes, 15 minutes, or 60 minutes, energy consumption and network load of battery-powered sensor devices, particularly those used in hard-to-reach areas, can be minimized. For example, a sensor device installed at a remote water treatment facility may transmit operating parameters of a centrifugal pump every 20 minutes to save energy, whereas high-performance industrial centrifugal pumps in a factory may be monitored continuously or every 10 minutes to respond early to abnormal conditions.

[0063] According to another preferred embodiment, detection is performed continuously and / or periodically, and / or every 500 ms, 250 ms, 100 ms, 50 ms, 25 ms, 5 ms, or 1 ms. Continuous recording means that the monitoring signal is recorded in real-time without interruption, which is particularly important in high-dynamic processes where load fluctuations are large and rapid changes in power consumption or vibration may occur in centrifugal pumps. On the other hand, periodic acquisition is performed at regular intervals, such as, for example, 100 ms or 500 ms, which may be sufficient when monitoring long-term or medium-term changes, such as a gradual increase in power consumption due to wear. The selection of a high sampling rate, for example, 1 ms, is particularly important in applications where centrifugal pump control must be adjusted appropriately by detecting rapid events, such as pressure surges or cavitation effects. Conversely, a low sampling rate of 500 ms or more is useful when only slowly changing operating conditions need to be checked, such as general energy consumption over a long period. By flexibly configuring the acquisition speed in this way, the sensor device can be optimized between high data resolution and energy-saving operation according to requirements, and data acquisition optimized for the application is ensured.

[0065] According to another preferred embodiment, the sensor device includes a mounting device for placement, in particular, on the top-hat rail of a control panel. The mounting device may consist of standardized snap-in or clamp fastening means, thereby allowing the sensor device to be mounted on the top-hat rail without additional tools or removed if necessary. A top-hat rail, also known as a DIN rail, is a standardized metal profile rail used to securely and space-efficiently mount electrical and electronic components, such as relays, control units, or sensor devices, in control panels and distribution boxes. A control panel is a housing that typically accommodates and protects electrical and electronic components, such as controllers, fuses, or sensors, in production facilities, water purification plants, or industrial pump systems. By attaching the sensor device to the top-hat rail within the control panel in this manner, the sensor device can be directly integrated into existing control systems, even retroactively, enabling a robust and professional installation, for example, for monitoring centrifugal pumps in water treatment facilities or industrial cooling circuits. This additional configuration facilitates a compact and compliant installation of the sensor device in industrial environments.

[0067] The objective of the present invention is also achieved by a centrifugal pump comprising the sensor device described above, wherein at least one signal input is connected to the centrifugal pump. A centrifugal pump generally refers to a fluid machine that primarily transports liquids using rotational motion and dynamic force. In a centrifugal pump, centrifugal force in radial flow is utilized for pumping in addition to tangential acceleration of the liquid, so such a pump is also called a centrifugal pump. The centrifugal pump can preferably be used in a building's hydraulic system or other applications, particularly as a wastewater pump.

[0069] During normal operation of the centrifugal pump, the motor housing of the centrifugal pump may be positioned on top of the pump housing, and an impeller driven by a shaft by the motor is provided within the pump housing to form a liquid transfer and centrifugal chamber. The motor housing may be connected to the pump housing in a fixed position or formed integrally. Additionally, the centrifugal pump and the motor may each have shafts, and these shafts may be connected to each other via couplings. Preferably, the shaft protrudes from the motor housing to the pump housing on the drive side, or is fixed to the drive side shaft of the impeller. Accordingly, the suction port for the fluid to be transferred is preferably positioned at the bottom of the pump housing. The motor housing and / or pump housing are preferably formed of metal, particularly cast iron or stainless steel, ceramic and / or plastic. The fluid or liquid preferably comprises other liquid media such as water or wastewater. The fluid or liquid may contain dirt, impurities, or various wastes, particularly feces, sediment, dust, sand, or solids such as small pieces of wood, weeds, fibers, or pieces of cloth, which may cause clogging within the pump housing.

[0071] The objective of the present invention is also achieved by a method for monitoring a centrifugal pump using a sensor device, said method,

[0072] A step of detecting at least one monitoring signal characterizing the state of a centrifugal pump at least once per second at at least one signal input of a sensor device, wherein the signal input includes a current transformer input, an analog input, and / or a digital input;

[0073] In particular, regarding the alarm status of the centrifugal pump, statistical evaluation of the centrifugal pump status and / or monitoring and prediction of the centrifugal pump status, the step of performing analysis and / or data reduction preprocessing on the at least one monitoring signal; and

[0074] The method includes the step of periodically transmitting at least one preprocessed monitoring signal to a cloud infrastructure configured for post-processing of the transmitted preprocessed monitoring signal via a low-power wide-area network (LPWAN) wireless interface of the sensor device at a frequency not exceeding once every 5 minutes.

[0076] According to a preferred additional embodiment, the method is,

[0077] It further includes a step of post-processing at least one pre-processed monitoring signal transmitted using cloud infrastructure.

[0079] These additional configurations enable more advanced analysis, long-term data evaluation, and / or intelligent prediction for centrifugal pump monitoring. Post-processing may include, for example, merging and comparing pre-processed monitoring signals transmitted over a long period to detect abnormal conditions indicating wear trends or imminent failure of the centrifugal pump. For instance, in a water treatment facility, the cloud can determine from transmitted pre-processed monitoring signals that the centrifugal pump's power consumption is gradually increasing, recognize this as an indication of increased bearing friction, and generate early maintenance recommendations. In industrial cooling systems, the cloud can compare historical data to detect that abnormal vibrations indicate cavitation or pipe blockage, enabling targeted action. Furthermore, the cloud can perform automatic pattern recognition using machine learning to predict future malfunctions and optimize preventive maintenance strategies. As another example, by integrating into a higher-level monitoring system that links various sensor devices within a production facility, correlations between centrifugal pump failures and other pump conditions can be identified. This cloud-based post-processing significantly improves the efficiency and reliability of centrifugal pump monitoring, while reducing the local computation and storage requirements of sensor devices.

[0081] According to another preferred embodiment, the analysis and / or data reduction preprocessing includes data compression and / or selection for at least one monitoring signal.

[0083] According to a preferred additional embodiment, the analysis and / or data reduction preprocessing comprises amplitude reduction and / or time reduction of at least one monitoring signal.

[0085] Additional embodiments and advantages of the above method are derived based on the sensor device described above. Unless a technically explicit order is required, the order of the method steps may be changed, and the order of steps described above is particularly preferred. Brief explanation of the drawing

[0087] Hereinafter, the present invention will be described in more detail according to a preferred embodiment with reference to the attached drawings. FIG. 1 illustrates a schematic cross-sectional view of a centrifugal pump for carrying out the proposed method according to a preferred embodiment of the present invention. FIG. 2 illustrates a schematic configuration including a centrifugal pump according to FIG. 1 and a sensor device for monitoring the centrifugal pump according to a preferred embodiment of the present invention. FIG. 3 illustrates a flowchart for carrying out the proposed method according to a preferred embodiment of the present invention. FIG. 4 illustrates a diagram showing the recording and evaluation of a monitoring signal for measuring current in a centrifugal pump according to a preferred embodiment of the present invention. FIG. 5 illustrates a diagram showing a trend analysis for monitoring the power consumption of a centrifugal pump according to a preferred embodiment of the present invention. FIG. 6 illustrates a diagram of continuously checking a threshold value at a signal input unit of a sensor device according to a preferred embodiment of the present invention. FIG. 7 illustrates another diagram of continuously checking a threshold value at the signal input section of a sensor device according to a preferred embodiment of the present invention. FIG. 8 illustrates a diagram showing the flow of additional information being transmitted to a cloud infrastructure located downstream of a sensor device according to a preferred embodiment of the present invention. Specific details for implementing the invention

[0088] FIG. 1 illustrates a schematic diagram of a centrifugal pump (1) for performing the method described below to monitor the centrifugal pump (1) according to a preferred embodiment of the present invention.

[0090] The centrifugal pump (1) is formed as a sewage pump and typically includes a pump housing (2), and in the drawing, a suction opening (3) is positioned at the bottom of the pump housing (2) as an inlet. Inside the pump housing (2), a rotor (4) extending in a vertical direction in the drawing is provided. The rotor (4) is part of an electric motor (5) positioned opposite the suction opening (3), and the electric motor (5) is only partially shown in FIG. 1. An impeller (6) is positioned toward the suction opening (3), and the impeller (6) is driven by the rotor (4) of the electric motor (5). Additionally, the centrifugal pump (1) includes a control system not shown in FIG. 1.

[0092] FIG. 2 illustrates a schematic configuration including a centrifugal pump (1) according to FIG. 1 and a sensor device (7) for monitoring the centrifugal pump (1) according to a preferred embodiment of the present invention. The sensor device (7) has its own housing separated from the centrifugal pump (1) and can be mounted on a top-hat rail of a control panel (not shown) via a fixing device (8). The sensor device (7) may be powered by a battery, but in this embodiment, it receives power from the control panel. The sensor device (7) includes a plurality of signal input units (9), which include a plurality of current transformer inputs, analog inputs, and digital inputs. Additionally, an analog output, at least one digital output, an RS485 interface, and a USB-C interface are provided.

[0094] One of the current transformer inputs is connected to a current transformer (10), which detects the current flowing on one phase of the power connection cable (11) of the centrifugal pump (1) and provides it as a monitoring signal characterizing the state of the centrifugal pump (1). The sensor device (7) also includes a processor-based data processing device (12) in the form of a microcontroller, which is shown only indirectly in FIG. 2, and a low-power wide area network (LPWAN) wireless interface (13). The data processing device (12) is configured to execute the method steps shown in FIG. 3, which is to perform the proposed method according to a preferred embodiment of the present invention.

[0096] In the first step (100), a monitoring signal characterizing at least one state of the centrifugal pump (1) is detected by a data processing device (12) at least every second at at least one signal input (9) of the sensor device (7). The monitoring signal may be the current consumption of the electric motor (5), the rotational speed of the centrifugal pump (1), etc., as previously described. Detection is performed continuously or periodically, preferably every 500, 250, 100, 50, 25, 5, or 1 ms.

[0098] In the second step (200), the data processing device (12) performs analysis and data reduction preprocessing on at least one monitoring signal in relation to the alarm status of the centrifugal pump (1), statistical evaluation of the state of the centrifugal pump (1), and / or monitoring and prediction of the state of the centrifugal pump (1). To this end, the analysis and data reduction preprocessing include data compression and selection for at least one monitoring signal.

[0100] An alarm state of the centrifugal pump (1) occurs, for example, when the sensor device (7) detects a critical operating deviation indicating a failure, overload, or imminent failure through analysis preprocessing. For example, a sudden increase in current consumption may indicate a blockage or bearing damage of the rotor shaft (5), and abnormal vibration may indicate cavitation or imbalance. If the temperature of the electric motor (5) exceeds a set limit, overheating may be detected, thereby triggering an alarm and generating a maintenance message. This alarm state may be transmitted to an upper monitoring system via a downstream cloud infrastructure (14) to enable an immediate response.

[0102] Statistical evaluation of the condition of the centrifugal pump involves identifying average values, trends, and deviations, for example, through long-term analysis of recorded monitoring signals. For example, a sensor device (7) can record and analyze average current consumption, speed, or temperature over several weeks to detect abnormal changes in operating conditions early. A typical application example is calculating the average vibration amplitude over several operating cycles to detect gradual changes due to mechanical wear. Additionally, trend analysis of energy consumption helps identify inefficiencies that cause greater load on the centrifugal pump (1), such as sediment in the piping. Such statistical evaluation forms the basis of optimized maintenance strategies and contributes to preventing unplanned downtime.

[0104] Condition monitoring is performed by continuously analyzing the monitoring signal of, for example, the centrifugal pump (1) to determine whether it is operating within the normal operating range. Various measurements, such as current consumption, speed, temperature, or vibration, are monitored in real time and can be compared with historical data. Condition prediction, or predictive maintenance, enables the prediction of future failures or maintenance needs to a significant extent using machine learning methods or predefined models based on recorded monitoring signals. For example, if current consumption increases under constant load conditions, it can be predicted that bearing damage is imminent and maintenance of the centrifugal pump (1) will be required within a few weeks. Similarly, through regular analysis of the vibration frequency spectrum, the occurrence of imbalance can be detected early before it leads to critical damage. This combination of condition monitoring and prediction enables efficient maintenance planning, reduces unplanned downtime, and extends the lifespan of the centrifugal pump (1).

[0106] In the third step (300), at least one pre-processed monitoring signal is periodically transmitted via the Low Power Wide Area Network (LPWAN) wireless interface (13) of the sensor device (7) to a cloud infrastructure (14) configured for post-processing of the transmitted pre-processed monitoring signal at a frequency not exceeding once every 5 minutes. Transmission is performed continuously or periodically and is not performed more frequently than 10, 15, 20, 30, or 60 minutes. The LPWAN wireless interface (13) is configured to transmit at least one pre-processed monitoring signal using the Long Range Wide Area Network (LoRaWAN) network protocol, the Long Term Evolution for Machines (LTE-M) network protocol, and / or the NarrowBand IoT network protocol. Due to the pre-processing performed in advance, the data transmission rate of at least one monitoring signal is 300 times, 500 times, 1000 times, or 2000 times greater than the data transmission rate of the transmitted pre-processed monitoring signal.

[0108] The transmission interval being performed at a frequency not exceeding once every 5 minutes means, for example, that data aggregation is performed within the sensor device (7) itself. For example, if multiple state changes, such as the on / off of the centrifugal pump (1), are detected within a transmission interval of 5 minutes, these state changes are counted, and up to a certain number, each with a unique timestamp, for example, up to 15 state changes within 5 minutes can be aggregated into a single message transmitted from the sensor device (7) to the cloud infrastructure (14). Accordingly, information is not provided in real-time (live), but information within a time range shorter than the transmission interval is not lost.

[0110] Finally, at least one transmitted preprocessed monitoring signal may be postprocessed, for example, at a control station (15) using cloud infrastructure (14). This postprocessing may include, for example, merging and comparing preprocessed monitoring signals transmitted over a long period to detect wear trends or abnormal conditions indicating an imminent failure of the centrifugal pump (1).

[0112] Additionally, preprocessing may be configured from the cloud infrastructure (14) and, for example, may be set to individually adjust thresholds for changes or events for each centrifugal pump (1). Furthermore, additional processing algorithms, such as filter algorithms, may be downloaded from the cloud infrastructure (14) to the sensor device (7) or data processing device (12) and selectively applied to individual monitoring signals.

[0114] FIG. 4 illustrates a diagram showing the recording and evaluation of a monitoring signal for current measurement in a centrifugal pump (1) according to a preferred embodiment of the present invention. The monitoring signal is provided in the form of current from a signal input unit (12) to detect the operating state of the centrifugal pump (1). When the measured current exceeds a set threshold value indicated as P1 in FIG. 4 or falls below a threshold value indicated as P2 in FIG. 4, the data processing unit (12) provides information through analysis preprocessing along with a timestamp, respectively. The average current value between P1 and P2 and the average value during continuous operation up to the next update point are displayed. The recorded monitoring signal may be permanently stored in a sensor device (7) so as to be maintained even in the event of a power outage. In another variation, the standard deviation for the ON period of the centrifugal pump (1), as well as the magnitude and duration of the start pulse, are also recorded and evaluated.

[0116] FIG. 5 illustrates a trend analysis diagram for monitoring the power consumption of a centrifugal pump (1) according to a preferred embodiment of the present invention. This analysis enables the early identification of potential problems of the centrifugal pump (1). A sensor device (7) or a data processing device (12) provides a counter for the number of times the centrifugal pump (1) has been started over the past 24 hours, including the operating cycle in progress at the time of evaluation. This counter is updated during each update cycle and is stored permanently so as to be maintained even in the event of a power outage. Alternatively, an alarm may be generated on a dashboard provided by the cloud infrastructure (14) based on a simple threshold comparison of an average value, such as 1.5 times the rated current, for example. In a later step, additional components of the cloud infrastructure (14) can perform a trend analysis on the reported average value to predict problems before an absolute threshold is exceeded.

[0118] FIG. 6 illustrates a diagram of continuously checking a threshold value at the signal input of a sensor device (7) for an adjustable ON / OFF detection threshold value according to a preferred embodiment of the present invention. When the measured monitoring signal exceeds the ON threshold value, an ON event is transmitted to the cloud infrastructure (14) along with a timestamp. Subsequently, when the measured monitoring signal falls below the OFF threshold value, an OFF event is transmitted to the cloud infrastructure (14) along with a timestamp. In addition, the time when the signal input (9) exceeds the threshold value, i.e., the ON time, is also transmitted to the cloud infrastructure (14) in two forms. While the analog input exceeds the ON threshold value, the elapsed time after exceeding the threshold value is transmitted every 15 minutes. When the analog input falls below the OFF threshold value, the time between the ON threshold value and the OFF threshold value is transmitted.

[0120] FIG. 7 illustrates another diagram in which a threshold value is continuously checked at a signal input (9) of a sensor device (7) for an adjustable ON / OFF detection threshold value according to a preferred embodiment of the present invention, wherein an alarm threshold value may be set. When the alarm threshold value is exceeded, an alarm message is transmitted to the cloud infrastructure (14) along with a timestamp. When the alarm threshold value is not exceeded, corresponding information including a timestamp indicating that the alarm is no longer active is transmitted to the cloud infrastructure (14).

[0122] FIG. 8 illustrates a diagram showing the flow of additional information being transmitted to a cloud infrastructure (14) located downstream of a sensor device (7) to detect the number of centrifugal pumps (1) operating simultaneously, according to a preferred embodiment of the present invention. When the centrifugal pump (1) exceeds a threshold value for ON detection, the counter is incremented by 1 and transmitted to the cloud infrastructure (14) along with a timestamp. When the value of the centrifugal pump (1) falls below the threshold value for ON detection, the counter is decremented by 1 and transmitted to the cloud infrastructure (14) along with a timestamp.

[0124] The embodiments described above are merely examples and may be modified and / or supplemented in various ways within the scope of the claims. Each feature described for a specific embodiment may be used independently or in combination with other features of other embodiments. Additionally, each feature described for an embodiment belonging to a specific category may be applied to embodiments belonging to other categories in a corresponding manner. Explanation of the symbols

[0126] 1: Centrifugal pump 2: Pump housing 3: Suction opening 4: Rotor 5: Electric motor 6: Impeller 7: Sensor device 8: Fixing device 9: Signal input section 19: Current transformer 11: Power connection cable 12: Data processing unit 13: LPWAN Wireless Interface 14: Cloud Infrastructure 100: A step of detecting at least one monitoring signal characterizing the state of the centrifugal pump (1) at least every second at at least one signal input part (9) of the sensor device (7). 200: A step of performing analysis and data reduction preprocessing on at least one monitoring signal in relation to the alarm status of the centrifugal pump (1), statistical evaluation of the status of the centrifugal pump (1) and / or monitoring and prediction of the status of the centrifugal pump (1). 300: Step of periodically transmitting at least one preprocessed monitoring signal to a cloud infrastructure (14) configured for post-processing of the transmitted preprocessed monitoring signal, through the low-power wide area network (LPWAN) wireless interface (13) of the sensor device (7) so as not to exceed once every 5 minutes.

Claims

Claim 1 A sensor device for monitoring a centrifugal pump, comprising a housing separate from the housing of the centrifugal pump, at least one signal input including a current transformer input, an analog input and / or a digital input, a low-power wide area network (LPWAN) wireless interface, and a processor-based data processing device, wherein the data processing device is configured to detect at least one monitoring signal characterizing the state of the centrifugal pump through the at least one signal input at least once per second, and is configured to perform analysis and data reduction preprocessing on the at least one monitoring signal in relation to at least one of an alarm state of the centrifugal pump, a statistical evaluation of the state of the centrifugal pump, and monitoring and prediction of the state of the centrifugal pump, and wherein the at least one preprocessed monitoring signal is transmitted at a frequency not exceeding once every five minutes through the LPWAN wireless interface to a cloud infrastructure configured for post-processing of at least one transmitted preprocessed monitoring signal. Claim 2 The sensor device according to claim 1, wherein the sensor device comprises at least one of two or three current transformer inputs, at least one analog output, at least one digital output, and at least one RS485 interface. Claim 3 A sensor device according to claim 1, wherein the analysis and data reduction type preprocessing comprises at least one of data compression and selection for at least one monitoring signal. Claim 4 A sensor device according to claim 1, wherein the analysis and data reduction type preprocessing comprises at least one of amplitude reduction and time reduction for at least one monitoring signal. Claim 5 A sensor device according to claim 1, wherein the data transmission rate of the at least one monitoring signal is 300 times, 500 times, 1000 times, or 2000 times or more than the data transmission rate of the at least one transmitted preprocessed monitoring signal. Claim 6 A sensor device according to claim 1, wherein at least one transmitted preprocessing monitoring signal is of Boolean data type. Claim 7 A sensor device according to claim 1, wherein the LPWAN wireless interface is configured to transmit the at least one preprocessing monitoring signal by at least one of the Long Range Wide Area Network (LoRaWAN) network protocol, the Long Term Evolution for Machines (LTE-M) network protocol, and the NarrowBand IoT network protocol. Claim 8 A sensor device according to claim 1, wherein the transmission is performed in at least one of continuously and periodically. Claim 9 A sensor device according to claim 1, wherein the transmission is performed not exceeding once every 10, 15, 20, 30, or 60 minutes. Claim 10 A sensor device according to claim 1, wherein the detection is performed in at least one of continuously and periodically. Claim 11 A sensor device according to claim 1, wherein the detection is performed at least once every 500 ms, 250 ms, 100 ms, 50 ms, 25 ms, 5 ms, or 1 ms. Claim 12 A sensor device according to claim 1, comprising a fixing device for placing the sensor device on a top hat rail or the top hat rail of a control panel. Claim 13 A centrifugal pump comprising a sensor device according to claim 1, wherein at least one signal input part is connected to the centrifugal pump. Claim 14 A method for monitoring a centrifugal pump using a sensor device having a housing separated from the housing of the centrifugal pump, comprising: a step of detecting at least one monitoring signal characterizing the state of the centrifugal pump at least once per second at at least one signal input of the sensor device, wherein the signal input includes at least one of a current transformer input, an analog input, and a digital input; a step of performing analysis and data reduction preprocessing on the at least one monitoring signal in relation to at least one of an alarm state of the centrifugal pump, a statistical evaluation of the state of the centrifugal pump, and monitoring and prediction of the state of the centrifugal pump; and a step of periodically transmitting the at least one preprocessed monitoring signal to a cloud infrastructure configured for post-processing of the transmitted preprocessed monitoring signal through a low-power wide-area network (LPWAN) wireless interface of the sensor device, such that the transmission does not exceed once every five minutes. Claim 15 A method according to claim 14, further comprising the step of post-processing at least one transmitted pre-processed monitoring signal using the cloud infrastructure. Claim 16 In claim 14, the above analysis and data reduction type preprocessing comprises at least one of data compression and selection for at least one monitoring signal. Claim 17 In claim 14, the above analysis and data reduction type preprocessing comprises at least one of amplitude reduction and time reduction for at least one monitoring signal.

Images