Pneumatic equipment and method for monitoring pneumatic equipment

Abstract

According to exemplary embodiments, a method for monitoring pneumatic equipment is provided. The method comprises: a step of collecting a data set of a plurality of operating variables of a pneumatic facility including a plurality of pneumatic components; and a step of generating a facility monitoring model, wherein the facility monitoring model is generated by clustering the data set of the plurality of operation variables on the basis of density.

Description

Pneumatic equipment and methods for monitoring pneumatic equipment

[0001] The present invention relates to pneumatic equipment and a method for monitoring pneumatic equipment. This application claims the benefit of Korean Application No. 10-2023-0179446, filed December 12, 2023, which is incorporated herein by reference in its entirety.

[0002] Unlike primary batteries, secondary batteries can be charged and discharged multiple times. They are widely used as a power source for various wireless devices, including handsets, laptops, and cordless vacuum cleaners. Recently, improved energy density and economies of scale have dramatically reduced the per-unit manufacturing cost of secondary batteries. Furthermore, as the range of battery electric vehicles (BEVs) has increased to match that of fuel-powered vehicles, the primary use of secondary batteries is shifting from mobile devices to mobility.

[0003] In the secondary battery manufacturing process, pneumatic equipment is used to handle and transport materials (such as battery cells or electrode semi-finished products). Pneumatic equipment is difficult to precisely control and has difficulties predicting failures in advance. Therefore, various studies are being conducted to predict failures in pneumatic equipment in advance, thereby preventing yield declines in secondary battery manufacturing and improving throughput.

[0004] The technical idea of ​​the present invention is to provide a pneumatic equipment and a method for monitoring the pneumatic equipment.

[0005] According to exemplary embodiments of the present invention for solving the above-described problem, a method for monitoring pneumatic equipment is provided. The method comprises the steps of collecting a data set of a plurality of operating variables of a pneumatic equipment including a plurality of pneumatic components; and generating a facility monitoring model, wherein the facility monitoring model is generated by clustering the data set of the plurality of operating variables based on density.

[0006] The above data set of the above multiple motion variables is clustered based on mean shift clustering.

[0007] The method further comprises the steps of collecting data points of the plurality of operating variables of the pneumatic equipment; and monitoring the pneumatic equipment based on the data points and the equipment monitoring model.

[0008] The step of monitoring the pneumatic equipment includes calculating a score indicating the condition of the pneumatic equipment based on the distance of the data points from normal clusters determined based on clustering of the data sets of the plurality of operating variables.

[0009] The method further includes a step of generating an alarm based on the score.

[0010] The method further comprises a step of preprocessing the data set of the plurality of operating variables.

[0011] The step of preprocessing the data set of the plurality of operating variables classifies the data set of the plurality of operating variables of the pneumatic equipment based on the operating state of the pneumatic equipment.

[0012] The data sets of the plurality of operating variables are classified into a normal data set of the plurality of operating variables and a failure data set of the plurality of operating variables.

[0013] The normal data set of the plurality of operating variables includes an optimal data set of the plurality of operating variables.

[0014] The above model is generated based on the above normal data set of the above plurality of operating variables.

[0015] The step of preprocessing the data set of the plurality of operating variables standardizes the data set of the plurality of operating variables of the pneumatic equipment.

[0016] The step of preprocessing the data set of the plurality of operating variables removes duplication and missingness of the data set of the plurality of operating variables of the pneumatic equipment.

[0017] The step of preprocessing the data set of the plurality of operating variables synchronizes the time of the data set of the plurality of operating variables of the pneumatic equipment.

[0018] According to exemplary embodiments, a pneumatic installation is provided. The installation comprises: a plurality of pneumatic components; sensors configured to detect a plurality of operating variables of the plurality of pneumatic components; and a processor configured to collect a data set of the plurality of operating variables, wherein the processor comprises a device monitoring model generated by clustering the data set of the plurality of operating variables.

[0019] The above facility monitoring model is generated by clustering the data set of the plurality of operating variables based on density.

[0020] The above data set of the above multiple motion variables is clustered based on mean shift clustering.

[0021] The processor is configured to monitor the pneumatic equipment based on data points of the plurality of operating variables of the pneumatic equipment and the equipment monitoring model.

[0022] The processor is configured to calculate distances of the data points from normal clusters determined based on clustering of the data set of the plurality of operating variables.

[0023] The processor is configured to produce a score indicating the condition of the pneumatic equipment based on distances of the data points from the normal clusters.

[0024] The processor is configured to generate an alarm based on the score.

[0025] A method for monitoring pneumatic equipment according to exemplary embodiments of the present invention can cluster normal data of a plurality of operating variables of a pneumatic equipment, and calculate distances of data points of a plurality of operating variables occurring in real time from the clusters determined accordingly. The method for monitoring pneumatic equipment according to exemplary embodiments can generate an alarm or control the operation of the pneumatic equipment based on a change in a score calculated from the distances, thereby providing a prediction of the status (e.g., normal and faulty) of the pneumatic equipment.

[0026] The effects that can be obtained from the exemplary embodiments of the present invention are not limited to the effects mentioned above, and other effects not mentioned can be clearly derived and understood by those skilled in the art to which the exemplary embodiments of the present disclosure pertain from the following description. In other words, unintended effects resulting from practicing the exemplary embodiments of the present disclosure can also be derived by those skilled in the art from the exemplary embodiments of the present disclosure.

[0027] Figure 1 is a block diagram showing a pneumatic installation according to exemplary embodiments.

[0028] Figure 2 is a flowchart illustrating a facility monitoring method according to exemplary embodiments.

[0029] Figures 3 to 8 are drawings for explaining examples of multiple operating variables measured by multiple sensors.

[0030] Figures 9 and 10 illustrate examples of data sets of multiple motion variables.

[0031] Figure 11 is a graph illustrating a facility monitoring method according to exemplary embodiments.

[0032] Figure 12 shows the change in scores according to experimental examples.

[0033] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings. Prior to this, it should be noted that the terms and words used in this specification and claims should not be construed as limited to their conventional or dictionary meanings. Based on the principle that the inventor can appropriately define the concepts of terms to best explain his or her invention, they should be interpreted in a way that aligns with the technical spirit of the present invention.

[0034] Accordingly, the embodiments described in this specification and the configurations illustrated in the drawings are merely the most preferred embodiments of the present invention and do not represent all of the technical ideas of the present invention. Therefore, it should be understood that there may be various equivalents and modified examples that can replace them at the time of filing this application.

[0035] In addition, when describing the present invention, if it is determined that a detailed description of a related known configuration or function may obscure the gist of the present invention, the detailed description is omitted.

[0036] Since the embodiments of the present invention are provided to more fully explain the present invention to those skilled in the art, the shapes and sizes of components in the drawings may be exaggerated, omitted, or schematically illustrated for clearer explanation. Accordingly, the sizes and proportions of each component do not fully reflect the actual sizes or proportions.

[0037]

[0038] (Embodiments 1 and 2)

[0039] FIG. 1 is a block diagram showing a pneumatic installation (100) according to exemplary embodiments.

[0040] Figure 2 is a flowchart illustrating a facility monitoring method according to exemplary embodiments.

[0041] Referring to FIGS. 1 and 2, at P110, a data set of multiple operating variables (V1, V2, V3) can be collected.

[0042] The pneumatic equipment (100) may be configured to perform, for example, a secondary battery manufacturing process. The pneumatic equipment (100) may be configured to perform, for example, any one of a mixing process, an electrode process, an assembly process, and an activation process. The pneumatic equipment (100) may include a plurality of pneumatic components (110), sensors (120), a controller (130), and a processor (140).

[0043] The electrode process may include a coating process, a roll pressing process, and a slitting process. In the coating process, an electrode slurry may be coated on an electrode sheet. The electrode slurry may include an active material, a conductive agent, a binder, and a solvent. The electrode slurry may be provided by dissolving the active material, the conductive agent, and the binder in a solvent. In the roll pressing process, the electrode sheet coated with the electrode slurry may be passed between pressurized rolls. The roll pressing process may flatten the surface of the electrode sheet and enhance the bonding strength between the active material and the current collector in the electrode sheet. The slitting process may separate the electrode sheet into a plurality of electrode sheets. The slitting process may be an optional process. Depending on the specifications of the electrode assembly of the final battery cell to be manufactured, the slitting process of the slitting device may be omitted. The electrode rolls completed in the slitting device may be processed by a winding device or a notching device. Accordingly, a stack-type electrode assembly or a cylindrical electrode assembly may be provided.

[0044] A notching process may be performed on an electrode sheet unwound from an electrode roll by a rewinder. In the notching process, electrode tabs may be formed on the uncoated portion of the electrode sheet. A V-shaped groove may be further formed on the electrode sheet. In the lamination process, individualized unit electrodes may be bonded to a separator. In the lamination process, the separator and the unit electrode may be heat treated to enhance the bonding strength between the separator and the unit electrode. The lamination process may provide a half-cell or a mono-cell. A bi-cell, etc. may also be provided by the lamination process. A half-cell may include a separator and an anode, or a separator and an anode. A mono-cell may include a cathode, a first separator, a cathode, and a second separator sequentially stacked. In the stacking process, at least one of the half-cell, mono-cell, and bi-cell may be repeatedly stacked. An electrode assembly may be provided by the stacking process. The electrode assembly may include a tape for securing the cathodes and the anodes. In the folding process, half-cells, mono-cells, and bi-cells may be wrapped around a separator. The packaging process may include inserting the electrode assembly into a case, injecting the electrolyte, and sealing the case.

[0045] The activation process may include charge / discharge, aging, degassing, performance testing, and Open Circuit Voltage (OCV)-based testing. During the activation process, a SEI film may form on the surface of the anode during the initial charge. The SEI film is a thin film that forms on the surface of the anode material when the battery cell is first charged after manufacturing. When the battery cell is charged, lithium ions within the battery cell migrate to the anode, and during this process, substances in the electrolyte undergo electrolysis for the first time, resulting in a chemical reaction that may form a Solid Electrolyte Interphase (SEI) film on the surface of the anode material. The SEI may be a type of separator. The SEI can prevent further decomposition reactions of the electrolyte during the migration of lithium ions from the anode to the anode for battery charging. During the aging process, the charged or discharged battery cell may be stored at room temperature for a predetermined period of time (for non-limiting examples, 30 minutes to 3 hours) to stabilize. The purpose of the aging process is to evenly distribute the electrolyte within the pouch cell, allowing it to permeate both the positive and negative electrodes. The aging process can improve lithium ion mobility. Gases may be generated within the battery cell during the charge / discharge and aging processes. The degassing process can remove these gases. Performance testing can include capacity testing and screening for defects. OCV-based testing can identify defective battery cells. Low-voltage defects in battery cells can be caused by metallic foreign matter located within the battery cell. For example, if the positive electrode of a battery cell contains metallic foreign matter, such as iron or copper, these foreign matters can grow into dendrites on the negative electrode. Dendrites can cause unwanted short circuits within the battery cell, which can lead to battery cell failure and fire.

[0046] Each of the plurality of pneumatic components (110) may be configured to operate based on pneumatic pressure or to induce and control pneumatic pressure applied to other elements. The plurality of pneumatic components (110) may include actuating devices such as pneumatic cylinders, rotary actuators, and air chucks, and directional control devices such as solenoid valves and mechanical valves. The solenoid valve can determine the forward and backward movement of the actuating components by changing the direction of compressed air. The plurality of pneumatic components (110) may further include electronic devices such as servo motors. The servo motors perform a similar function to pneumatic cylinders, but can provide more precise control than pneumatic cylinders.

[0047] A plurality of sensors (120) may be configured to detect the operation of a plurality of pneumatic components (110). The plurality of sensors (120) may be configured to determine a plurality of operating variables (V1, V2, V3) of the plurality of pneumatic components (110).

[0048] The controller (130) may be configured to control a plurality of pneumatic components (110). The controller (130) may be configured to receive signals representing a plurality of operating variables (V1, V2, V3) from a plurality of sensors (120). The controller (130) may be configured to determine a status of the plurality of pneumatic components (110). For example, the controller (130) may be configured to collect data including an error code indicating whether some of the plurality of pneumatic components (110) are in a normal operating state or a fault state. The controller (130) may be configured to transmit a signal representing the plurality of operating variables (V1, V2, V3) to the processor (140).

[0049] Here, the controller (130) may be a Programmable Logic Controller (PLC). A PLC is a special type of microprocessor-based controller that uses programmable memory to store commands and implement functions such as logic, sequencing, timing, counting, and arithmetic to control machines and processes. PLCs are easy to operate and program.

[0050] The controller (130) may include a power supply, a central processing unit (CPU), an input interface, an output interface, a communication interface, and memory devices. The power supply may be configured to supply power to other elements of the controller (130), such as the CPU, the input interface, the output interface, the communication interface, and the memory devices, for the operation of the controller (130). The memory devices may include a read-only memory (ROM) configured to store a system program, such as an operating system, and a random access memory (RAM) configured to store data, such as user programs and status information of input and output devices, timers, counters, and other internal device values. The CPU may be configured to control communication between modules that implement logic and convert input signals into output operation signals. The CPU may operate based on the system program and the user program stored in the memory devices. The CPU may be configured to write or read inspection data and measurement data to the data area of ​​the memory devices based on the system program and the user program. Conditions or data of industrial devices and production processes may be transmitted to the CPU through the input module. The results processed by the CPU can be transmitted to the actuator via the output module. The communication interface can be configured to relay the transmission and reception of data between the controller (130) and the processor (140).

[0051] However, the controller (130) is not limited thereto, and may include any one of a simple controller, a complex processor such as a microprocessor, a CPU, a GPU, a processor configured by software, dedicated hardware, and firmware. The controller (130) may be implemented by, for example, a general-purpose computer or application-specific hardware such as a DSP (Digital Signal Processor), an FPGA (Field Programmable Gate Array), and an ASIC (Application Specific Integrated Circuit).

[0052] Figures 3 to 8 are drawings for explaining examples of multiple operating variables (V1, V2, V3) measured by multiple sensors (120). More specifically, Figures 3 to 8 show the reciprocation of a pneumatic cylinder (112) according to a control signal applied to a solenoid valve (111) from a controller (130).

[0053] Referring to FIGS. 3 to 8, the pneumatic cylinder (112) may include first and second contact sensors (112S1, 112S2). As in FIGS. 3 and 8, when the pistol (112P) of the pneumatic cylinder (112) is in a rearward position, the first contact sensor (112S1) may be in an on state, and the second contact sensor (112S2) may be in an off state. As in FIGS. 5 and 6, when the pistol (112P) of the pneumatic cylinder (112) is in a forward position, the first contact sensor (112S1) may be in an off state, and the second contact sensor (112S2) may be in an on state. As shown in FIGS. 4 and 7, when the pistol (112P) of the pneumatic cylinder (112) moves between the forward position and the backward position, the first contact sensor (112S1) and the second contact sensor (112S2) may be in an off state.

[0054] In Fig. 3, the solenoid valve (111) can be switched from an off state to an on state. Accordingly, as in Fig. 4, due to the switching of the solenoid valve (111) to an on state, the pistol (112P) of the pneumatic cylinder (112) can be moved forward, and the first contact sensor (112S1) can be switched to an off state. The forward reaction time may be the time required between the situation (Configuration) of Fig. 3 and the situation of Fig. 4. More specifically, the forward reaction time may be the time taken for the pistol (112P) of the pneumatic cylinder (112) to start moving forward after the solenoid valve (111) is switched to an on state.

[0055] The point in time at which the solenoid valve (111) is switched to the on state can be determined by the point in time at which the control signal of the controller (130) controlling the solenoid valve (111) is transmitted to the solenoid valve (111). Accordingly, the forward reaction time can be the difference between the point in time at which the control signal of the controller (130) controlling the solenoid valve (111) is transmitted to the solenoid valve (111) and the point in time at which the first sensor (111S1) is switched from the on state to the off state.

[0056] In Fig. 5, the pistol (112P) of the pneumatic cylinder (112) may reach a forward position, and accordingly, the second contact sensor (112S2) may be switched to an on state. The forward movement time may be the time required between the situations of Fig. 4 and Fig. 5. More specifically, the forward movement time may be the time required until the pistol (112P) moves to the forward position after the movement of the pistol (112P) begins. The forward movement time may be the difference between the time when the first sensor (111S1) switches from an on state to an off state and the time when the second sensor (111S2) switches from an off state to an on state.

[0057] In Fig. 6, the solenoid valve (111) can be switched from an on state to an off state. In Fig. 7, due to the switching of the solenoid valve (111) to an off state, the pistol (112P) of the pneumatic cylinder (112) moves backward, and the second contact sensor (112S2) can be switched to an off state. The backward reaction time may be the time required between the situations of Fig. 6 and Fig. 7. More specifically, the backward reaction time may be the time taken for the pistol (112P) of the pneumatic cylinder (112) to start moving backward after the solenoid valve (111) is switched to an off state.

[0058] The point in time at which the solenoid valve (111) is switched to the off state can be determined by the point in time at which the control signal of the controller (130) controlling the solenoid valve (111) is transmitted to the solenoid valve (111). Accordingly, the reverse reaction time can be the difference between the point in time at which the control signal of the controller (130) controlling the solenoid valve (111) is transmitted to the solenoid valve (111) and the point in time at which the second contact sensor (112S2) is switched from the on state to the off state.

[0059] In Fig. 8, the pistol (112P) of the pneumatic cylinder (112) may reach the rearward position, and accordingly, the first contact sensor (112S1) may be switched to the on state. The rearward movement time may be the time required between the situations of Fig. 7 and Fig. 8. More specifically, the rearward movement time may be the time required until the pistol (112P) moves to the rearward position after the movement of the pistol (112P) begins. The rearward movement time may be the difference between the time when the second sensor (111S2) switches from the on state to the off state and the time when the first sensor (111S1) switches from the off state to the on state.

[0060] In FIGS. 3 to 8, the collection of data sets of four multiple operating variables from two pneumatic components has been described, but the dimension of the collected data may vary depending on the number of sensors applied to the pneumatic components. For example, if a flow sensor and a pressure sensor are provided in either the solenoid valve (111) or the pneumatic cylinder (112), the flow sensor may detect the flow rate of either the solenoid valve (111) or the pneumatic cylinder (112), and the pressure sensor may detect the pneumatic pressure of either the solenoid valve (111) or the pneumatic cylinder (112). Accordingly, two additional data sets of multiple operating variables may be collected.

[0061] As another example, the pneumatic components (110) may include a vacuum chuck and a solenoid valve configured to control the pressure applied to the vacuum chuck. A plurality of sensors (120) may be configured to derive an on-duty response time, an on-duty, an off-duty, and an off-duty response time of the vacuum chuck from the vacuum chuck and the solenoid valve. The on-duty and the off-duty of the vacuum chuck may be determined based on the pressure sensor. In the on-duty of the vacuum chuck, a high pneumatic pressure may be applied to the vacuum chuck, and in the off-duty of the vacuum chuck, a low pneumatic pressure may be applied to the vacuum chuck. The on-duty response time and the off-duty response time of the vacuum chuck may be the time required from the time when the controller (130) transmits a control signal of the solenoid valve to the time when the pressure applied to the vacuum chuck changes.

[0062] As another example, the pneumatic components (110) may include a servo motor, and the plurality of operating variables (V1, V2, V3) may include a load factor of the servo motor. Here, the load factor of the servo motor may represent a ratio of the output of the servo motor to the maximum output of the servo motor. The load factor of the servo motor may be determined based on either the driving current or the driving power of the servo motor.

[0063] A plurality of sensors (120) may be configured to transmit a plurality of operating variables (V1, V2, V3) to a processor (140). The processor (140) may be configured to collect a data set of the plurality of operating variables (V1, V2, V3). The plurality of operating variables (V1, V2, V3) may be collected by the processor (140) via the controller (130). As another example, the processor (140) may be configured to load a data set including a plurality of operating variables (V1, V2, V3) stored in a server such as an MES, a data warehouse, and a network attached storage.

[0064] Each of the data points of the data set of the plurality of operating variables (V1, V2, V3) may be a multidimensional quantity. The dimension of each of the data points of the data set of the plurality of operating variables (V1, V2, V3) may correspond to different plurality of operating variables (V1, V2, V3). For example, in order to collect the data set of the plurality of operating variables (V1, V2, V3), M plurality of operating variables may be extracted from N parts, and in this case, each of the data points of the data set of the plurality of operating variables (V1, V2, V3) may have a value of dimension M. Here, N and M may each be integers. M may be greater than N, but is not limited thereto.

[0065] Figures 9 and 10 illustrate examples of data sets of multiple operating variables (V1, V2, V3). More specifically, Figure 9 illustrates the forward reaction time and forward movement time of an electrode transport device, and Figure 10 illustrates the forward reaction time and forward movement time of an electrode tab cutting device. In Figures 9 and 10, the horizontal axis represents the collection time of data points, and the vertical axis is the time axis of the forward reaction time and forward movement time. In Figures 9 and 10, the circular points represent the forward movement time, and the triangular points represent the forward reaction time.

[0066] Referring to FIGS. 9 and 10, it was confirmed that the scatter plot (i.e., the layering phenomenon) of the electrode tab cutting device having a relatively large number of parts was greater than that of the electrode transfer device having a relatively small number of parts.

[0067] Referring back to FIGS. 1 and 2, at P120, a data set of a plurality of operating variables (V1, V2, V3) may be preprocessed. The processor (140) may be configured to preprocess the data set of the plurality of operating variables (V1, V2, V3). The preprocessing of the data set of the plurality of operating variables (V1, V2, V3) may include classification of the data set of the plurality of operating variables (V1, V2, V3), time synchronization of the data set of the plurality of operating variables (V1, V2, V3), normalization of the data set of the plurality of operating variables (V1, V2, V3), and removal of duplication and missingness of the data set of the plurality of operating variables (V1, V2, V3).

[0068] Classification of the data set of the plurality of operating variables (V1, V2, V3) may include determining the state of the pneumatic equipment (100) at the time the data set of the plurality of operating variables (V1, V2, V3) was collected. According to exemplary embodiments, the processor (140) may be configured to generate a equipment monitoring model based on a portion of the data set of the plurality of operating variables (V1, V2, V3) selected based on the classification of the data set of the plurality of operating variables (V1, V2, V3).

[0069] For example, the operating status of the pneumatic equipment (100) may include normal and faulty. If an error code occurs in the controller (130) for some of the plurality of pneumatic components (110), the operating status of the pneumatic equipment (100) may be faulty. Conversely, if no error code occurs in the controller (130), the operating status of the pneumatic equipment (100) may be normal. Here, the error code may be recorded and collected by the controller (130) when the value of the sensors (120) exceeds the upper threshold or falls below the lower threshold. As a non-limiting example, the error code may be transmitted by the controller (130) to a server such as an MES (Manufacturing Execution System) and stored in the MES. The data set of multiple operating variables (V1, V2, V3) may include a portion collected when the operating status of the secondary battery manufacturing facility is normal, and the data set of multiple operating variables (V1, V2, V3) may include a portion collected when the operating status of the secondary battery manufacturing facility is faulty.

[0070] Among the data sets of the plurality of operating variables (V1, V2, V3), a portion collected when the operating status of the secondary battery manufacturing facility is normal may be referred to as a normal data set of the plurality of operating variables (V1, V2, V3). Among the data sets of the plurality of operating variables (V1, V2, V3), a portion collected when the operating status of the secondary battery manufacturing facility is faulty may be referred to as a fault data set of the plurality of operating variables (V1, V2, V3). In some cases, an optimal data set of the plurality of operating variables (V1, V2, V3) may be selected from the normal data sets of the plurality of operating variables (V1, V2, V3). The normal data set of the plurality of operating variables (V1, V2, V3) may include the optimal data set of the plurality of operating variables (V1, V2, V3). By selecting an optimal data set of multiple operating variables (V1, V2, V3), a portion deteriorated due to the operation of multiple pneumatic components (110) can be excluded from the normal data set of multiple operating variables (V1, V2, V3).

[0071] The time resolution of at least some of the plurality of sensors (120) may be on the order of milliseconds, and thus, each of the plurality of sensors (120) may detect and store the plurality of operating variables (V1, V2, V3) at different timings. According to exemplary embodiments, the data points of each of the plurality of operating variables (V1, V2, V3) may be divided into units of set time intervals (e.g., 1 second), and a representative value of the divided sections may be selected, thereby compensating for the difference in the collection timing of the plurality of operating variables (V1, V2, V3). Here, the representative value of the plurality of operating variables (V1, V2, V3) may be any one of the mean, standard deviation, median, maximum, and minimum values ​​within the sections of the plurality of operating variables (V1, V2, V3).

[0072] Normalization of a data set of multiple operating variables (V1, V2, V3) may be standard normalization. Normalization of a data set of multiple operating variables (V1, V2, V3) may include dividing the multiple operating variables (V1, V2, V3) by the absolute value of the maximum value of each of the multiple operating variables (V1, V2, V3). Each of the multiple operating variables (V1, V2, V3) has a different value depending on the operating pattern of a related one of the multiple pneumatic components (110). Accordingly, some of the multiple operating variables (V1, V2, V3) may have a relatively larger absolute value than other some of the multiple operating variables (V1, V2, V3).

[0073] The method for monitoring the condition of the equipment described below relies on the distance between clusters of a data set consisting of a plurality of operating variables (V1, V2, V3) and newly measured data points, and if the data set of the plurality of operating variables (V1, V2, V3) and the newly measured data points of the plurality of operating variables (V1, V2, V3) are not normalized, unintended weights are applied to the plurality of operating variables (V1, V2, V3).

[0074] For example, in the absence of normalization, an air cylinder having a relatively long forward movement time among the plurality of operating variables (V1, V2, V3) due to differences in driving patterns has greater importance in monitoring the equipment status than an air cylinder having a relatively short forward movement time among the plurality of operating variables (V1, V2, V3). According to exemplary embodiments, the importance of each of the plurality of operating variables (V1, V2, V3) can be equalized through normalization of the data set of the plurality of operating variables (V1, V2, V3).

[0075] In addition, duplication and missingness of the data set of the plurality of operating variables (V1, V2, V3) may distort the distribution of the data set composed of the plurality of operating variables (V1, V2, V3). According to exemplary embodiments, by removing duplication and missingness of the data set of the plurality of operating variables (V1, V2, V3), the distortion of the distribution of the data set used for generating the facility monitoring model can be prevented, and the reliability of the pneumatic facility (100) and the method for monitoring the pneumatic facility (100) can be improved.

[0076] Next, a facility monitoring model can be generated in P130. The facility monitoring model can be included in the processor (140). The facility monitoring model can be generated by the processor (140). According to exemplary embodiments, the facility monitoring model can be generated based on a data set of a plurality of data operating variables (V1, V2, V3). According to exemplary embodiments, the facility monitoring model can be generated based on a normal data set of a plurality of data operating variables (V1, V2, V3). According to exemplary embodiments, the facility monitoring model can be generated based on an optimal data set of a plurality of data operating variables (V1, V2, V3).

[0077] Generating a facility monitoring model may include classifying any one of a data set of a plurality of operating variables (V1, V2, V3), a normal data set of the plurality of operating variables (V1, V2, V3), and an optimal data set of the plurality of operating variables (V1, V2, V3) into a plurality of clusters based on density. According to exemplary embodiments, the facility monitoring model of the processor (140) may be generated by mean shift clustering.

[0078] Mean-shift clustering can involve calculating the probability density function of the centroid of a cluster using Kernel Density Estimation (KDE) and shifting the centroid in a direction that increases the probability density function. The centroid shift can be repeated for a specified number of iterations or terminated when no clusters shift. KDE can be defined according to the following equation.

[0079]

[0080] Here, K represents the kernel function, x represents the random variable value, xi represents the data points of the normal data set of multiple operating variables (V1, V2, V3), and h represents the bandwidth. Setting the bandwidth is necessary for mean-shift clustering. Mean-shift clustering can provide more rigorous and flexible clustering because it does not require settings for the data set shape or the number of clusters.

[0081] Examples of kernels may include a uniform kernel, a triangular kernel, a biweight kernel, a Laplacian kernel, a Cauchy kernel, a rectangular kernel, an Epanechnikov kernel, a normal kernel, and a Gaussian kernel. A kernel is a positive-valued function that can satisfy normalization and symmetry conditions.

[0082] According to exemplary embodiments, the facility monitoring model of the processor (140) may be generated by either K-Means Clustering or DBSCAN (Density Based Spatial Clustering of Applications with Noise).

[0083] K-means clustering establishes cluster centers and clusters data points based on the distance between the centers and the data points. The cluster centers are then moved to the average point of the data, and the data points are re-clustered based on the shifted cluster centers. In K-means clustering, the number of cluster centers is preset during the facility monitoring model design, and the iteration can be repeated until the centers no longer move.

[0084] DBSCAN involves finding core points that satisfy the minimum number of data points within an epsilon radius, and connecting the core points if they exist among neighboring points. Here, neighboring points are other data points located within the epsilon radius. DBSCAN requires the epsilon radius and the minimum number of data points to be set. In DBSCAN, the number of neighboring points of a border point is less than the minimum number of data points, but the neighboring points of a border point may include core points. In DBSCAN, the number of neighboring points of a noise point is less than the minimum number of data points, and the neighboring points of a noise point do not include core points.

[0085] Figure 11 is a graph illustrating a facility monitoring method according to exemplary embodiments.

[0086] Referring to FIGS. 1, 2 and 11, at P140, a pneumatic installation (100) can be monitored. The pneumatic installation (100) can be monitored by a processor (140). At P130, a data set of a plurality of operating variables (V1, V2, V3) can be assigned to a plurality of clusters (G1, G2, G3, G4), one of a normal data set of the plurality of operating variables (V1, V2, V3) and an optimal data set of the plurality of operating variables (V1, V2, V3).

[0087] Hereinafter, an example will be described in which multiple clusters (G1, G2, G3, G4) are determined by assignment of normal data sets of multiple operating variables (V1, V2, V3). In this case, the multiple clusters (G1, G2, G3, G4) may represent different phases of a pneumatic equipment (100) in a normal operating state.

[0088] For example, one of the plurality of clusters (G1, G2, G3, G4) may be a non-operating cluster, another of the plurality of clusters (G1, G2, G3, G4) may be an initial operating cluster, and another of the plurality of clusters (G1, G2, G3, G4) may be a main operating cluster. The non-operating cluster may be a sub-cluster of a normal data set of the plurality of operating variables (V1, V2, V3) resulting from a phase in which the pneumatic equipment (100) is not operating. The initial operating cluster may be a sub-cluster of a normal data set of the plurality of operating variables (V1, V2, V3) resulting from a phase before each element (e.g., the plurality of pneumatic components (100)) enters a normal state after the pneumatic equipment (100) starts operating. The main operating cluster may be a sub-cluster of a normal data set of a plurality of operating variables (V1, V2, V3) resulting from a phase after the pneumatic equipment (100) starts operating and each element (e.g., a plurality of pneumatic components (100)) enters a normal state.

[0089] The pneumatic equipment (100) can be monitored by a processor (140). The processor (140) can be configured to monitor the pneumatic equipment (100) based on data points of operating variables (V1, V2, V3) of the pneumatic equipment (100) and a equipment monitoring model.

[0090] Monitoring of the pneumatic equipment (100) may include calculating distances of new data points (NDP) collected from a plurality of sensors (120) from center points (C1, C2, C3, C4) of each of the clusters (G1, G2, G3, G4) of the normal data set. The processor (140) may be configured to calculate distances of the data points (NDP) from center points (C1, C2, C3, C4) of each of the clusters (G1, G2, G3, G4) of the normal data set.

[0091] In Fig. 11, two-dimensional distances are shown for convenience of illustration, but the distances from clusters (G1, G2, G3, G4) may be distances calculated in the same dimension as the data points of the data set of multiple operating variables (V1, V2, V3). For example, if the data points of the data set of multiple operating variables (V1, V2, V3) are M-dimensional, the distances from clusters of the normal data set may be distances calculated in the M-dimensional space.

[0092] According to exemplary embodiments, the monitoring of the pneumatic equipment (100) may further include calculating a score based on distances of data points from clusters of a data set of a plurality of operating variables (V1, V2, V3). The score may be calculated by the processor (140). The score may be calculated based on a representative value of distances between clusters (e.g., center points of clusters) and data points of the data set of the plurality of operating variables (V1, V2, V3). Here, the representative value may be any one of an average, a standard deviation, a median, a maximum, and a minimum of distances between clusters (e.g., center points of clusters) and data points of the data set of the plurality of operating variables (V1, V2, V3).

[0093] The method for monitoring equipment may further include a step of generating an alarm based on the score. According to exemplary embodiments, the processor (140) may be configured to generate a command for controlling the operation of the pneumatic equipment (100) or to provide an alarm to an operator based on a comparison of the score with a threshold. For example, the processor (140) may be configured to generate a command for controlling the operation of the pneumatic equipment (100) or to provide an alarm to an operator when the score changes below or above a threshold.

[0094] Figure 12 shows the change in scores according to experimental examples.

[0095] Referring to FIGS. 1, 2, 11 and 12, monitoring of the pneumatic installation (100) at P140 may include monitoring a score (SC) of data points (NDP) generated from sensors (120) of the pneumatic installation (100). In FIG. 12, new data points (NDP) are collected in real time, and thus, a score (SC) can be calculated in near real time. In FIG. 12, the representative distance (RD) and the score (SC) can be calculated based on four plurality of operating variables (Va, Vb, Vc, Vd). The plurality of operating variables (Va, Vb, Vc, Vd) are examples of the plurality of operating variables (V1, V2, V3) of FIG. 1.

[0096] At a first time point (t1), an alarm may be generated due to a change in the score. At the first time point, a stratification phenomenon of multiple operating variables (Va, Vb, Vc, Vd) was confirmed. The stratification phenomenon of multiple operating variables (Va, Vb, Vc, Vd) means that multiple operating variables (Va, Vb, Vc, Vd) deviate from a normal state. Subsequently, at a second time point (t2) adjacent to the first time point (t1), a failure of the pneumatic equipment (100) was confirmed. At the second time point (t2), as a failure of the pneumatic equipment (100) occurs, a peak of the representative distance (RD) may occur, and a valley of the score (SC) may occur. In the experimental example of Fig. 12, it was confirmed that a method for monitoring pneumatic equipment according to exemplary embodiments provides a prediction of a failure of the pneumatic equipment (100).

[0097]

[0098] The present invention has been described in more detail through drawings and examples. However, the configurations described in the drawings or examples described in this specification are merely embodiments of the present invention and do not represent all of the technical ideas of the present invention. Therefore, it should be understood that various equivalents and modified examples may exist as of the time of this application.

Claims

1. A step of collecting a data set of multiple operating variables of a pneumatic installation including multiple pneumatic components; and Including the steps of creating a facility monitoring model, A method for monitoring pneumatic equipment, characterized in that the equipment monitoring model is generated by clustering the data set of the plurality of operating variables based on density.

2. In paragraph 1, A method for monitoring pneumatic equipment, characterized in that the data set of the plurality of operating variables is clustered based on mean shift clustering.

3. In paragraph 1, A step of collecting data points of the plurality of operating variables of the pneumatic equipment; and A method for monitoring a pneumatic installation, further comprising the step of monitoring the pneumatic installation based on the data points and the installation monitoring model.

4. In paragraph 3, A method for monitoring pneumatic equipment, wherein the step of monitoring the pneumatic equipment comprises calculating a score representing the condition of the pneumatic equipment based on the distance of the data points from normal clusters determined based on clustering of the data sets of the plurality of operating variables.

5. In paragraph 4, A method for monitoring pneumatic equipment further comprising the step of generating an alarm based on the score.

6. In paragraph 1, A method for monitoring pneumatic equipment further comprising the step of preprocessing said data set of said plurality of operating variables.

7. In paragraph 6, A method for monitoring pneumatic equipment, characterized in that the step of preprocessing the data set of the plurality of operating variables of the pneumatic equipment classifies the data set of the plurality of operating variables of the pneumatic equipment based on the operating state of the pneumatic equipment.

8. In paragraph 7, A method for monitoring pneumatic equipment, characterized in that the data sets of the plurality of operating variables are classified into a normal data set of the plurality of operating variables and a failure data set of the plurality of operating variables.

9. In paragraph 8, A method for monitoring a pneumatic installation, characterized in that the normal data set of the plurality of operating variables includes an optimal data set of the plurality of operating variables.

10. In paragraph 8, A method for monitoring pneumatic equipment, wherein the model is generated based on the normal data set of the plurality of operating variables.

11. In paragraph 6, A method for monitoring pneumatic equipment, characterized in that the step of preprocessing the data set of the plurality of operating variables of the pneumatic equipment comprises standardizing the data set of the plurality of operating variables of the pneumatic equipment.

12. In paragraph 6, A method for monitoring pneumatic equipment, characterized in that the step of preprocessing the data set of the plurality of operating variables removes duplication and missingness of the data set of the plurality of operating variables of the pneumatic equipment.

13. In paragraph 6, A method for monitoring a pneumatic facility, characterized in that the step of preprocessing the data set of the plurality of operating variables comprises synchronizing the time of the data set of the plurality of operating variables of the pneumatic facility.

14. Multiple pneumatic components; Sensors configured to detect a plurality of operating variables of the plurality of pneumatic components; and A processor configured to collect a data set of the above plurality of operating variables, A pneumatic installation, characterized in that the processor includes a facility monitoring model generated by clustering the data set of the plurality of operating variables.

15. In paragraph 14, A pneumatic facility, characterized in that the above facility monitoring model is generated by clustering the data set of the plurality of operating variables based on density.

16. In paragraph 14, A pneumatic installation, characterized in that the data set of the plurality of operating variables is clustered based on mean shift clustering.

17. In paragraph 14, A pneumatic installation, characterized in that the processor is configured to monitor the pneumatic installation based on data points of the plurality of operating variables of the pneumatic installation and the installation monitoring model.

18. In paragraph 17, A pneumatic installation, characterized in that the processor is configured to calculate distances of the data points from normal clusters determined based on clustering of the data set of the plurality of operating variables.

19. In paragraph 18, A pneumatic installation, characterized in that the processor is configured to produce a score representing the condition of the pneumatic installation based on distances of the data points from the normal clusters.

20. In paragraph 19, A pneumatic installation, wherein the processor is configured to generate an alarm based on the score.

Images