KR20250054951A - A System for a Predictive Maintenance of a Motor Operating Valve Based on a IoT Sensor and an AI Algorithm - Google Patents

Abstract

The present invention relates to a predictive maintenance system for a motor-driven valve based on an IoT sensor and an AI algorithm. The predictive maintenance system for a motor-driven valve based on an IoT sensor and an AI algorithm includes an IoT sensor module (10) including a data acquisition module (11) for acquiring detection information from at least one sensor detecting an operating state of a motor (M) or a valve (V), an abnormality detection unit (12) for detecting an abnormality by comparing the acquired data with a reference value, and a first communication means (13) for transmitting the acquired data and the detected abnormality information by wire or wirelessly; and a diagnosis server including a data analysis module (22) for analyzing data and abnormality information transmitted from the IoT sensor module (10), an operation analysis module (23) for analyzing the operating state according to the analysis result of the data analysis module (22), an AI fault prediction module (24) for predicting a fault according to the analysis result of the operation analysis module (23), and a diagnosis module (25) for diagnosing a state according to the prediction result of the AI combined prediction module (24).

Description

{A System for a Predictive Maintenance of a Motor Operating Valve Based on a IoT Sensor and an AI Algorithm}

The present invention relates to a predictive maintenance system for a motor-driven valve based on an IoT sensor and an AI algorithm, and more particularly, to a predictive maintenance system for a motor-driven valve based on an IoT sensor and an AI algorithm, which can diagnose the operating status of a motor-driven valve by applying an AI algorithm to detection information transmitted in real time from an IoT sensor.

In the case where a motor-driven valve used to control the flow of fluid in a nuclear power plant, chemical plant, large-scale piping system or similar facility malfunctions, it may lead to a major accident due to the characteristics of the facility. Therefore, in order to prevent the failure of the motor-driven valve in advance, it is necessary to detect the operating status in real time and diagnose whether there is a possibility of malfunction. Regarding the diagnosis of such a motor-driven valve, Patent Publication No. 10-2012-0035994 discloses a real-time performance diagnosis method of a motor-driven valve that can monitor the operating status of the motor-driven valve and the prevention of damage to valve and actuator parts in real time using the stem friction coefficient of the actuator, and a diagnosis system using the same. In addition, Patent Publication No. 10-2004-0083131 discloses a device and method for monitoring the performance of a motor-driven valve. A motor-driven valve is composed of a plurality of parts, and the status of each part can affect the operating status of the motor-driven valve. Therefore, the status of each part must be detected and diagnosed, and the status of the motor-driven valve must be diagnosed based on this. However, prior art does not disclose such technology.

The present invention is intended to solve the problems of prior art and has the following objectives.

Prior art 1: Patent publication number 10-2012-0035994 (Korea Hydro & Nuclear Power Co., Ltd., published on April 17, 2012) Real-time performance diagnosis method of motor-driven valve and diagnosis system using the same Prior art 2: Patent publication number 10-2004-0083131 (Korea Advanced Institute of Science and Technology, published on October 1, 2004) Method for calculating stem force of motor-driven valve using electric power signal and performance diagnosis method

The purpose of the present invention is to provide a predictive maintenance system for a motor-driven valve based on an IoT sensor and an AI algorithm that enables real-time diagnosis of the overall operating status of a motor-driven valve by detecting the operating status of each component through a sensor module composed of various sensors.

According to a preferred embodiment of the present invention, a predictive maintenance system for a motor-driven valve based on an IoT sensor and an AI algorithm includes an IoT sensor module comprising a data acquisition module for acquiring detection information from at least one sensor detecting an operating state of a motor or a valve, an abnormality detection unit for detecting an abnormality by comparing the acquired data with a reference value, and a first communication means for transmitting the acquired data and the detected abnormality information by wire or wirelessly; and a diagnostic server comprising a data analysis module for analyzing data and abnormality information transmitted from the IoT sensor module, an operation analysis module for analyzing the operating state according to an analysis result of the data analysis module, an AI fault prediction module for predicting a fault according to the analysis result of the operation analysis module, and a diagnosis module for diagnosing a state according to a prediction result of the AI combined prediction module.

According to another suitable embodiment of the present invention, the IoT sensor module receives detection information detected by a sensor that detects the operating status of a motor, valve, and gear module; and voltage information, current information, or torque information of the motor.

According to another suitable embodiment of the present invention, the IoT sensor module further includes a reference range module for determining a reference value of acquired data; a deviation detection module for calculating a deviation level by which the data deviates from the reference range; and a risk calculation module for calculating a risk level based on the deviation level.

According to another suitable embodiment of the present invention, the diagnostic server further includes an operation criterion separation module for separating the motor-driven valves according to operation criteria and a correlation variable determination module for determining a correlation variable for each separated module.

The predictive maintenance system of a motor-driven valve based on an IoT sensor and an AI algorithm according to the present invention enables remote diagnosis of the motor-driven valve in real time. The predictive maintenance system according to the present invention enables diagnosis of various types of failures or malfunctions by detecting the status of each component constituting the motor-driven valve. The diagnostic system according to the present invention applies an AI algorithm to predict the possibility of failure in advance and enables predictive maintenance of the motor-driven valve. The predictive maintenance system according to the present invention may include various types of sensors, and the present invention is not limited thereby.

FIG. 1 illustrates an embodiment of a predictive maintenance system for a motor-driven valve based on an IoT sensor and an AI algorithm according to the present invention.
FIG. 2 illustrates an embodiment of a detection structure for a diagnostic system according to the present invention.
FIG. 3 illustrates an embodiment of a structure for predicting failure in a diagnostic system according to the present invention.
Figure 4 illustrates an embodiment of the operating structure of a diagnostic system according to the present invention.
FIG. 5 illustrates an embodiment of a predictive maintenance method for a motor-driven valve by a diagnostic system according to the present invention.

Below, the present invention is described in detail with reference to embodiments shown in the attached drawings, but the embodiments are for a clear understanding of the present invention and the present invention is not limited thereto. In the description below, components having the same drawing symbols in different drawings have similar functions and therefore will not be described repeatedly unless necessary for the understanding of the invention, and well-known components will be briefly described or omitted, but they should not be understood as being excluded from the embodiments of the present invention.

FIG. 1 illustrates an embodiment of a predictive maintenance system for a motor-driven valve based on an IoT sensor and an AI algorithm according to the present invention.

Referring to FIG. 1, a predictive maintenance system for a motor-driven valve using an IoT sensor and an AI algorithm includes an IoT sensor module (10) including a data acquisition module (11) for acquiring detection information from at least one sensor detecting an operating state of a motor (M) or a valve (V), an abnormal state detection unit (12) for detecting an abnormal state by comparing the acquired data with a reference value, and a first communication means (13) for transmitting the acquired data and the detected abnormal state information by wire or wirelessly; and a diagnostic server including a data analysis module (22) for analyzing data and abnormal state information transmitted from the IoT sensor module (10), an operation analysis module (23) for analyzing the operating state according to the analysis result of the data analysis module (22), an AI fault prediction module (24) for predicting a fault according to the analysis result of the operation analysis module (23), and a diagnosis module (25) for diagnosing a state according to the prediction result of the AI combined prediction module (24).

The predictive maintenance system may be composed of at least one sensor arranged in a motor-driven valve; an IoT sensor module (10) that acquires and processes information detected by the at least one sensor; and a diagnostic server (20) that analyzes detection data transmitted from the IoT sensor module (10) to predict whether a failure has occurred. At least one sensor may be an IoT sensor and may have a structure capable of short-distance communication with the IoT sensor module (10). The motor-driven valve includes a motor (M); a gear module rotated by the motor (M); a stem rotated by the gear module; and a valve (V) opened and closed by the stem, and an IoT sensor may be installed in at least one component of the motor-driven valve. The operating state of the motor (M) or the valve (V) may be detected by the IoT sensor, and detection data acquired by the IoT sensor may be transmitted to the data acquisition module (11) of the IoT sensor module (10). The data acquisition module (11) can convert the transmitted detection data into transmittable data, and the data acquisition module (11) can transmit the transmitted detection data to the abnormality detection module (12). The abnormality detection module (12) can include a reference value, and the reference value can be detection data detected by the IoT sensor when the motor-driven valve operates normally. The reference value can be a value detected by the IoT sensor in the initial installed state of the motor-driven valve or an average value of the detection data detected by the IoT sensor during the operation of the motor-driven valve. The reference value and the detection data can be compared by the abnormality detection module (12), and whether the detection data is repeatedly detected for a certain period of time while exceeding a predetermined range compared to the reference value can be detected. If such an abnormal state is detected, it can be stored and transmitted to the diagnostic server (20). The detection data and abnormal state information converted by the data acquisition module (11) can be transmitted to the second communication means (21) of the diagnostic server (20) through the first communication means (13).

The diagnostic server (20) may have a function of diagnosing the status of the motor-driven valve by analyzing information data transmitted from the IoT sensor module (10). Specifically, the transmitted detection data may be analyzed by the data analysis module (22). The data analysis module (22) may analyze the trend, average value, peak value, interconnectivity, or similar characteristics of the detection data. The analysis result may be transmitted to the operation analysis module (23), and the operation status of the motor-driven valve may be analyzed by the operation analysis module (23). The analysis of the operation status includes analysis of the operation status of each component constituting the motor-driven valve. According to the analysis result of the operation analysis module (23), a defect of each component may be predicted by the AI (Artificial Intelligence) defect prediction module (24). The AI fault prediction module (24) may include an artificial intelligence algorithm such as a RNN (Recurrent Neural Network) algorithm or an LSTM (Long Short Term Memory Models) algorithm, and the AI fault prediction module (24) may apply the artificial intelligence algorithm to predict a fault of a component of a motor-driven valve. Then, the result of such fault prediction may be transmitted to the diagnosis module (25), and the diagnosis module (25) may calculate a failure probability of each component according to the result of the fault prediction. Then, each component may be diagnosed based on the calculated failure probability, and the overall status of the motor-driven valve may be diagnosed based on this. Then, the motor-driven valve may be precisely diagnosed, or the component may be repaired or the component may be replaced according to the diagnosis result. Depending on the diagnosis result, various necessary measures may be taken, and the present invention is not limited thereby.

FIG. 2 illustrates an embodiment of a detection structure for a diagnostic system according to the present invention.

When detecting FIG. 2, the IoT sensor module (10) receives detection information detected by a sensor group (S1, S2, S3, S4) that detects the operating states of the motor (M), the valve (V), and the gear module (G1, G2); and voltage information, current information, or torque information of the motor (M). The motor-driven valve includes a motor (M); a connecting gear module (G1) such as a pinion gear or a helical gear connected to the motor (M); an operating gear module (G2) such as a worm gear or a worm wheel gear connected to and operated by the connecting gear module (G1); a limit switch or a torque switch (TS) for closing and opening, and a stem or valve (V). IoT sensors (S1, S2, S3, S4) for detecting the operating states of the motor (M), the gear module (G1, G2), the valve (V), and the stem may be installed. Each IoT sensor (S1, S2, S3, S4) may include a vibration sensor, a noise sensor, a displacement sensor, or a torque sensor, and the IoT sensor module (10) may receive voltage or current information applied to the motor (M). Each IoT sensor (S1, S2, S3, S4) may detect noise, vibration, displacement, torque, current, or voltage generated during operation in real time or by a detection start signal and transmit the detected noise, vibration, displacement, torque, current, or voltage to the IoT sensor module (10). Each IoT sensor (S1, S2, S3, S4) may communicate with the IoT sensor module (10) by a short-distance communication method such as Bluetooth communication, and the IoT sensor module (10) may be connected to a diagnostic server (20) so as to be able to communicate with it directly or through a repeater (MT). A risk assessment unit (RG) may be installed in an IoT sensor module (10) or a repeater (MT), and the IoT sensor module (10) or the repeater (MT) may have a structure capable of remote communication with a diagnostic server (20). The risk assessment unit (RG) may compare a pre-stored reference value with a measured value, and determine that the measured value is at a risk level if the measured value shows a difference of, for example, 20 to 50% compared to the reference value. In addition, the risk may be determined based on the degree of deviation from the reference value. The risk determined in this manner may be transmitted to the diagnostic server (20), and the diagnostic server (20) may take appropriate measures according to the transmitted risk. The risk may be determined in various ways, and the present invention is not limited thereby.

FIG. 3 illustrates an embodiment of a structure for predicting failure in a diagnostic system according to the present invention.

Referring to FIG. 3, the IoT sensor module (10) further includes a reference range module (14) for determining a reference value of acquired data; a deviation detection module (15) for calculating a deviation level at which data deviates from the reference range; and a risk calculation module (16) for calculating a risk level based on the deviation level. The reference range module (14) has a function of determining a reference value of a parameter measured by each IoT sensor. The reference value may be determined by the basic specifications of the motor-driven valve or may be a value measured in an initial normal operating state. For example, the reference value may be an average value measured for each parameter for a certain period of time in an initial normal operating state. Alternatively, an average value of the remaining values excluding the upper 20% and the lower 20% from the parameter values measured in an operating state of the motor-driven valve for a certain period of time may be the reference value. In this manner, when the reference value is determined by the reference range module (14), whether a parameter corresponding to the deviation range is measured may be detected by the deviation detection module (15). The deviation detection module (15) can determine that a parameter value measured by an IoT sensor in real time is, for example, 30% or more or 30% or less than a reference value, as falling within a deviation range. The deviation detection module (15) can detect the occurrence of the deviation range, duration, or deviation range and transmit them to the risk calculation module (16). The risk calculation module (16) can calculate the risk level according to the deviation level detected by the deviation detection module (15). In addition, when the risk level is above a certain level, the risk calculation module (16) can generate an alarm to notify the necessity of immediate inspection, repair, replacement, or similar immediate measures. The parameter value measured by each IoT sensor can be transmitted to the diagnostic server (20) through the IoT sensor module (10) together with the calculated risk level. The diagnostic server (20) includes a parameter analysis module (22), and the parameter analysis module (22) can analyze noise, vibration, displacement, current, voltage, or similar parameters to determine whether the device is operating in a normal state. In addition, the parameter analysis module (22) can analyze the correlation between different parameters. The analysis result by the parameter analysis module (22) can be transmitted to the failure probability calculation module (26). The failure probability calculation module (26) can calculate the failure probability from the analysis result of the parameters, and can calculate the failure probability based on the abnormal state value appearing in each parameter value, the number of times the abnormal state value appears, the change trend between parameter values having a correlation, or the change trend of similar parameters. The failure probability calculation module (26) can calculate the failure probability based on the number of times a value above or below a predetermined range appears. For example, the failure probability can be calculated based on the number of times a difference value of 70% or more or a difference value of 30% or less is detected compared to the average value. If the failure probability exceeds a predetermined level, such as 50% or more, appropriate measures can be taken accordingly. In addition, detectable values of the subsequently determined parameters can be predicted by the parameter value prediction module (27). An artificial intelligence algorithm can be applied to predict such parameter values. An artificial intelligence algorithm can be applied based on the fluctuation trends of various parameters and various types of failure states to predict parameter values. For example, the period or number of times a specific range value of a parameter corresponding to an abnormal state appears can be predicted. Or, the correlation between different parameters can be predicted. Thereafter, the predicted value and the measured value can be compared by the measurement/prediction comparison module (28), and thereby the accuracy of the prediction according to the artificial intelligence algorithm can be verified. Thereafter, the artificial intelligence algorithm is applied to predict the parameter values again, and the measured value and the predicted value can be compared by the measurement/prediction comparison module (28). The fluctuation of the parameter values by the artificial intelligence algorithm can be predicted in various ways, and the present invention is not limited thereby.

Figure 4 illustrates an embodiment of the operating structure of a diagnostic system according to the present invention.

Referring to FIG. 4, the diagnostic server (20) further includes an operation criterion separation module (44a) that separates the motor-driven valve according to the operation criterion, and a correlation variable determination module (44b) that determines a correlation variable for each separated module. Detection information of each IoT sensor acquired from the IoT sensor module (10) can be transmitted to the diagnostic server (20), and correlation between different parameters can be detected by the correlation detection module (42) through the variable change trend module (41) of the diagnostic server (20). In addition, a criterion range can be set by the criterion range setting module (43), and the criterion range can be an average value of the measured values for each parameter as described above. The modules of the motor-driven valve can be separated based on the operation by the operation criterion module separation module (44a). And parameters that can determine the operation status of each module can be determined by the module correlation variable determination module (44b). Such module-related variables can be reference parameters for judging the operating status of the module, and for example, a motor-driven valve can be separated into a motor module, a first gear module, a second gear module, a gear valve linkage module, and a valve module. And at least one module correlation variable can be determined for each module by a module correlation variable determination module (44b). When the module correlation variable is determined in this way, an artificial intelligence algorithm can be applied to each module by an AI algorithm change prediction module (45) so that parameter values for each module can be predicted. When the parameter value is predicted, the measured value and the predicted value can be compared by a measurement/prediction comparison module (46). And through this process, whether or not there is a failure can be predicted and diagnosed by a failure prediction diagnosis module (47). In order to verify the diagnosis process, a module and module correlation variables can be selected by a variable/module selection module (48). And the correlation variable selected by a variable designation module (49) can be designated and transmitted to an IoT sensor module (10). Afterwards, parameters according to the IoT sensor determined by the IoT sensor module (10) can be tracked according to a predetermined cycle. In this way, the accuracy of the prediction by the artificial intelligence algorithm can be verified. The prediction verification by the artificial intelligence algorithm can be performed in various ways, and the present invention is not limited thereby.

FIG. 5 illustrates an embodiment of a predictive maintenance method for a motor-driven valve by a diagnostic system according to the present invention.

Referring to FIG. 5, a predictive maintenance method of a motor-driven valve includes a step of arranging IoT sensors in each of a plurality of modules constituting the motor-driven valve (P51); a step of measuring at least one operation-related variable related to the operation of a plurality of modules of the motor-driven valve (P52); a step of setting a reference value of each operation-related variable (P53); a step of separating each module based on the operation of the motor-driven valve (P54); a step of detecting a correlation between the measured operation-related variables (P55); a step of determining whether a deviation value from the reference value is measured (P56); a step of determining module-related variables for each module separated based on the operation criteria (P57); a step of applying an artificial intelligence algorithm to predict a change in the determined module-related variable for each module (P58); a step of comparing the predicted change value with the measured value (P59); a step of determining whether a difference occurs between the change value and the measured value (P60); a step of predicting whether each module has a failure (P61); and a step of predictively diagnosing the failure (P62). and a step (P63) in which necessary actions are taken based on the predictive diagnosis of a failure.

If a deviation value out of the reference value is not measured (NO) in the process of determining the deviation value (P56), it corresponds to a normal operating state and operation-related variables can be continuously measured (P52). In contrast, if a deviation value is measured (YES), there is a possibility that the operating state is defective. In addition, the predicted value and the measured value are compared (P59), it is determined whether a difference has occurred (P60), and if it is determined that there is a difference (YES), the process of applying the artificial intelligence algorithm again to predict the change can proceed (P58). In contrast, if the difference value is below a predetermined range (NO), whether each module is faulty can be predicted (P61). Predictive maintenance of a motor-driven valve can be performed in various ways and the present invention is not limited thereby.

Although the present invention has been described in detail above with reference to the presented embodiments, those skilled in the art will be able to make various modifications and variations without departing from the technical spirit of the present invention by referring to the presented embodiments. The present invention is not limited by such modifications and variations, but is only limited by the claims attached below.

10: IoT sensor module 20: Diagnostic server
11: Data Acquisition Module 12: Anomaly Detection Unit
13: First communication means 14: Reference range module
15: Deviation Detection Module 16: Risk Calculation Module
22: Data Analysis Module 23: Operational Analysis Module
24: AI Defect Prediction Module 25: Diagnosis Module
44a: Operation Criteria Separation Module 44b: Correlation Variable Determination Module

Claims (4)

An IoT sensor module (10) comprising a data acquisition module (11) for acquiring detection information from at least one sensor that detects the operating status of a motor (M) or a valve (V), an abnormality detection unit (12) for detecting an abnormality by comparing the acquired data with a reference value, and a first communication means (13) for transmitting the acquired data and the detected abnormality information by wire or wirelessly; and
A predictive maintenance system for a motor-driven valve based on an IoT sensor and AI algorithm, comprising a diagnostic server comprising a data analysis module (22) that analyzes data and abnormal status information transmitted from an IoT sensor module (10), an operation analysis module (23) that analyzes an operation status based on the analysis results of the data analysis module (22), an AI fault prediction module (24) that predicts a fault based on the analysis results of the operation analysis module (23), and a diagnostic module (25) that diagnoses a status based on the prediction results of the AI combined prediction module (24). In claim 1, the IoT sensor module (10) is characterized in that it receives detection information detected by sensors (S1, S2, S3, S4) that detect the operating status of the motor (M), the valve (V), and the gear module (G1, G2); and voltage information, current information, or torque information of the motor (M). A predictive maintenance system for a motor-driven valve based on an IoT sensor and an AI algorithm. In claim 1, the IoT sensor module (10) further includes a reference range module (14) for determining a reference value of acquired data; a deviation detection module (15) for calculating a deviation level at which data deviates from the reference range; and a risk calculation module (16) for calculating a risk level based on the deviation level. A predictive maintenance system for a motor-driven valve based on an IoT sensor and an AI algorithm. In claim 1, a motor-driven valve predictive maintenance system based on IoT sensors and AI algorithms, wherein the diagnostic server (20) further includes an operation criterion separation module (44a) that separates the motor-driven valves according to operation criteria and a correlation variable determination module (44b) that determines correlation variables for each separated module.