Magnetization device degradation diagnosis system and magnetization device

Abstract

This invention provides a magnetization device degradation diagnosis system that can detect failures in the magnetization device in advance without affecting the energization conditions to the magnetization yoke, thereby preventing a decrease in productivity due to sudden failures of the magnetization device. [Solution] The system comprises a magnetizing yoke 1 in which a magnetizing coil 5 is wound around an iron core 4, a magnetizing power supply 2 connected to the magnetizing coil 5, a sensor 3 installed on the magnetizing yoke 1, and a signal processing device 6 connected to the sensor 3. The sensor 3 detects a signal caused by sound or vibration generated when current is supplied from the magnetizing power supply 2 to the magnetizing coil 6, and the signal processing device 6 diagnoses the deterioration state of the magnetizing yoke 1 based on the output of the sensor 3.

Description

Technical Field

[0001] The present disclosure relates to a deterioration diagnosis system for a magnetizing device and a magnetizing device.

Background Art

[0002] In the motor manufacturing process, magnetizing devices that pass an instantaneous large current through a magnetizing coil to magnetize a target object are widely used. Since a large current is repeatedly passed through the magnetizing coil, thermal deterioration and mechanical deterioration may eventually lead to dielectric breakdown and result in failure. Since the deterioration of the magnetizing yoke composed of an iron core and a magnetizing coil progresses internally, failures often occur suddenly, leading to reduced productivity due to production line shutdowns. To prevent this, it is necessary to detect failures of the magnetizing yoke in advance.

[0003] To solve this problem, there is disclosed a method that includes a magnetizing yoke having a magnetizing coil wound around an iron core, a search coil wound around the iron core, and a control unit that measures the induced voltage of the search coil when the magnetizing coil is energized, wherein the control unit predicts the service life of the magnetizing yoke based on the similarity between the measured waveform of the induced voltage measured for each magnetization and a reference waveform (for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, in Patent Document 1, the magnetic field generated by the induced voltage of the search coil is in the direction of canceling out the magnetization magnetic field, so the magnetization magnetic field decreases, and greater energy is required to sufficiently magnetize the magnet. Due to the increased energy requirement, there are concerns that the load during energization will be greater and the lifespan will be shorter compared to a magnetization yoke without a search coil.

[0006] This disclosure provides technology to solve the above-mentioned problems, and aims to provide a magnetizing device degradation diagnosis system and magnetizing device that can detect failures of the magnetizing device in advance without affecting the energizing conditions to the magnetizing yoke, and prevent a decrease in productivity due to sudden failures of the magnetizing device. [Means for solving the problem]

[0007] The magnetization device degradation diagnosis system of the present disclosure comprises a magnetization yoke in which a magnetization coil is wound around an iron core, a magnetization power supply connected to the magnetization coil, a sensor installed on the magnetization yoke, and a signal processing device connected to the sensor, wherein the sensor detects a signal caused by sound or vibration generated when current is supplied from the magnetization power supply to the magnetization coil, and the signal processing device diagnoses the degradation state of the magnetization yoke based on the output of the sensor. The magnetization device of this disclosure comprises a magnetization yoke in which a magnetization coil is wound around an iron core, a magnetization power supply connected to the magnetization coil, a sensor installed on the magnetization yoke for detecting a signal caused by sound or vibration generated when current is supplied from the magnetization power supply to the magnetization coil, and a signal processing device connected to the sensor. [Effects of the Invention]

[0008] The magnetization device degradation diagnosis system of this disclosure provides a system that can detect failures in the magnetization device in advance without affecting the energizing conditions to the magnetization yoke, thereby preventing a decrease in productivity due to sudden failures of the magnetization device. The magnetizing device of this disclosure provides a magnetizing device that can detect a failure in the magnetizing device in advance without affecting the energizing conditions to the magnetizing yoke, thereby preventing a decrease in productivity due to a sudden failure of the magnetizing device. [Brief explanation of the drawing]

[0009] [Figure 1] This is a schematic diagram showing the configuration of the magnetization device degradation diagnosis system according to Embodiment 1. [Figure 2] This is a schematic diagram showing the configuration of the magnetization yoke in the degradation diagnosis system of the magnetization device according to Embodiment 1. [Figure 3] This is a cross-sectional view showing the structure of the magnetization yoke of the magnetization device degradation diagnosis system according to Embodiment 2. [Figure 4] This is a longitudinal cross-sectional view showing the structure of the magnetization yoke of the magnetization device degradation diagnosis system according to Embodiment 3. [Figure 5] This is a schematic diagram showing the configuration of the magnetization device degradation diagnosis system according to Embodiment 4. [Figure 6] Figure 6A is an explanatory diagram of a sensor fixing jig for installing the sensor of the magnetization device deterioration diagnosis system according to Embodiment 4 onto the magnetization yoke. Figure 6B is an explanatory diagram of the installation of the sensor fixing jig for installing the sensor of the magnetization device deterioration diagnosis system according to Embodiment 4 onto the magnetization yoke. [Figure 7] This is an example of the sensor output of the degradation diagnosis system for the magnetization device according to Embodiment 5. [Figure 8] This is a diagram showing the configuration of the learning device for the magnetization device degradation diagnosis system according to Embodiment 5. [Figure 9] This is a conceptual diagram of the neural network model used in the learning process of the learning device of the magnetization device degradation diagnosis system according to Embodiment 5. [Figure 10] This is a flowchart relating to the learning process of the learning device of the magnetization device degradation diagnosis system according to Embodiment 5. [Figure 11] This is a diagram showing the configuration of the inference device for the degradation diagnosis system of the magnetization device according to Embodiment 5. [Figure 12]This is a flowchart related to inference processing of an inference device of a deterioration diagnosis system for a magnetizing device according to Embodiment 5. [Figure 13] This is an example of a sensor output of the deterioration diagnosis system for a magnetizing device according to Embodiment 6. [Figure 14] This is another example of a sensor output of the deterioration diagnosis system for a magnetizing device according to Embodiment 6. [Figure 15] This is a flowchart related to operation of the deterioration diagnosis system for a magnetizing device according to Embodiment 7. [Figure 16] This is a hardware configuration example of the signal processing device, the learning device, and the inference device according to Embodiment 1, Embodiment 4, and Embodiment 5. [Modes for Carrying Out the Invention]

[0010] Embodiment 1. Embodiment 1 includes a magnetizing yoke having a magnetizing coil wound around an iron core, a magnetizing power source connected to the magnetizing coil, a sensor disposed on the magnetizing yoke, and a signal processing device connected to the sensor, the sensor detects a signal caused by sound or vibration generated when electric current is supplied from the magnetizing power source to the magnetizing coil, and the signal processing device relates to a deterioration diagnosis system for a magnetizing device that diagnoses a deterioration state of the magnetizing yoke based on an output of the sensor.

[0011] Hereinafter, the deterioration diagnosis system for a magnetizing device according to Embodiment 1 will be described with reference to each drawing. In each drawing, identical or corresponding parts are denoted by identical reference numerals.

[0012] First, the overall configuration of the deterioration diagnosis system 100 for a magnetizing device will be described based on FIG. 1 and FIG. 2. FIG. 1 is a schematic diagram showing the configuration of the deterioration diagnosis system 100 for a magnetizing device according to Embodiment 1. FIG. 2 is a schematic diagram showing the structure of the magnetizing yoke 1 of the deterioration diagnosis system 100 for a magnetizing device. A deterioration diagnosis system 100 for a magnetizing device according to Embodiment 1 includes a magnetizing yoke 1, a magnetizing power supply 2 connected to the magnetizing yoke 1, a sensor 3 installed on and normally fixed to the magnetizing yoke 1, and a signal processing device 6 connected to the sensor 3.

[0013] The magnetizing yoke 1 is composed of an iron core 4 and a magnetizing coil 5 wound around the iron core 4, and the magnetizing coil 5 is connected to the magnetizing power supply 2. By releasing the electric charge accumulated in the capacitor of the magnetizing power supply 2 and inputting a pulsed voltage to the magnetizing coil 5, a large current is caused to flow instantaneously through the magnetizing coil 5, and a strong magnetic field generated by the magnetizing coil 5 is applied to a magnetizing target (not illustrated), thereby magnetizing the magnetizing target.

[0014] On the other hand, the sensor 3 is fixed to the magnetizing yoke 1, and detects a signal resulting from sound or vibration generated when the magnetizing coil 5 is energized. The purpose of the deterioration diagnosis system 100 for the present magnetizing device is to diagnose deterioration of the magnetizing yoke 1. When the magnetizing yoke 1 deteriorates, the sound and vibration generated during energization change, and for example, deformation and cracks occur due to deterioration of the magnetizing coil 5, and as a result, elastic waves (ultrasonic waves) are generated.

[0015] Therefore, the type of the sensor 3 is not limited to, for example, a microphone and a piezoelectric sensor that directly detect sound or vibration generated from the magnetizing yoke 1 in a normal state. A configuration using any of an AE (acoustic emission) sensor that detects elastic waves generated when deformation and cracks occur in the magnetizing yoke 1, an acceleration sensor, or a strain gauge may also be adopted. Note that in a case where the installation environment of the magnetizing device 10 is such that ambient sound and vibration are dominant over the output of the sensor 3, it is appropriate to use, for example, an AE sensor instead of a sound or vibration sensor. Further, the number of the sensors 3 is not limited to one as shown in FIG. 1. For example, a configuration in which a plurality of sensors are collectively fixed at one location, or fixed one by one at a plurality of locations may be adopted. When a plurality of sensors are used, a plurality of types of sensors may be used at the same time.

[0016] When describing the hardware configuration of the magnetization device degradation diagnosis system 100 excluding the diagnostic processing function, the magnetization yoke 1, which consists of an iron core 4 and a magnetization coil 5 wound thereon, the magnetization power supply 2, the sensor 3, and the signal processing device 6 are collectively referred to as the magnetization device 10.

[0017] Next, we will describe the deterioration diagnosis of the magnetization yoke 1 using the deterioration diagnosis system 100 of the magnetization device according to Embodiment 1. In the magnetization device degradation diagnosis system 100, a sensor 3 fixed to the magnetization yoke 1 detects a signal caused by sound or vibration generated when current is supplied to the magnetization coil 5. The output from sensor 3 is recorded in signal processing unit 6 each time power is applied. Based on the recorded sensor output value, signal processing unit 6 can diagnose the deterioration of magnetization device 10.

[0018] Furthermore, by fixing the sensor 3 to the existing magnetization yoke 1, it becomes possible to construct a magnetization device deterioration diagnosis system 100. This allows for diagnosis not only when introducing a new magnetization device, but also when assessing the deterioration status of existing equipment.

[0019] As described above, the magnetization device degradation diagnosis system of Embodiment 1 can detect failures in the magnetization device in advance without affecting the energizing conditions to the magnetization yoke, thereby realizing a magnetization device degradation diagnosis system that can prevent a decrease in productivity due to sudden failures of the magnetization device.

[0020] Embodiment 2. Embodiment 2 involves fixing the sensor to the radial outer surface of the magnetization yoke at a position where the distance from the magnetization coil is minimized.

[0021] The following describes the degradation diagnosis system for the magnetization device of Embodiment 2, focusing on the differences from Embodiment 1, based on Figure 3. Figures 1 and 2 of Embodiment 1 will be referenced as appropriate. In the drawings of Embodiment 2, parts that are the same as or corresponding to those in Embodiment 1 are denoted by the same reference numerals. For the purpose of distinguishing this from Embodiment 1, the magnetization device 20 and the magnetization device deterioration diagnosis system 200 are used.

[0022] Figure 3 is a cross-sectional view showing the structure of the magnetization yoke 1 in the magnetization device deterioration diagnosis system 200 according to Embodiment 2. The magnetization device deterioration diagnosis system 200 of Embodiment 2 includes a magnetization yoke 1 in which a magnetization coil 5 is wound around an iron core 4, a magnetization power supply 2 connected to the magnetization yoke 1, a sensor 3 fixed to the magnetization yoke 1, and a signal processing device 6 connected to the sensor 3.

[0023] A feature of the magnetization device deterioration diagnosis system 200 of this second embodiment is that, when the sensor 3 is fixed to the radial outer surface of the magnetization yoke 1, the sensor 3 is fixed at a position where the distance L1 from the magnetization coil 5 to the sensor 3 is shortest.

[0024] Regarding the position of the magnetization coil 5, when the distance to the radial outer surface is smaller than the distance to the axial end surface, radial installation can suppress the distance between the magnetization coil and the sensor, allowing for the acquisition of good data. In particular, when a cooling structure (using water and coolant) is located on the axial outside of the magnetization coil 5, radial installation allows for the acquisition of more sensitive and better data.

[0025] Note that the number of magnetic poles of the magnetized yoke 1 is not limited to the number of poles shown in Figure 3 (8 poles). Furthermore, the shape of the magnetization yoke 1 is not limited to the circular shape shown in Figure 3. For example, it may have a polygonal shape other than a circle. Furthermore, the contact surface of the sensor 3 with the magnetization yoke 1 is not limited to the flat surface shown in Figure 3; it may also be a curved surface. Furthermore, the combination of the outer circumferential shape of the magnetizing yoke 1 and the contact surface shape of the sensor 3 is not limited to the combination of flat and curved surfaces shown in Figure 3. For example, both the outer circumferential surface of the magnetizing yoke 1 and the contact surface of the sensor 3 may be flat or curved, or the outer circumferential surface of the magnetizing yoke 1 may be flat and the contact surface of the sensor 3 may be curved.

[0026] The type of sensor 3 is not limited to those that directly detect sound or vibration. For example, a configuration using an AE sensor that detects elastic waves, or an acceleration sensor or strain gauge, may be used as a sensor to detect deformation and cracks. Furthermore, the number of sensors 3 is not limited to the one shown in Figure 3. For example, multiple sensors may be fixed together in one location, or one sensor may be fixed in multiple locations. When using multiple sensors, multiple types of sensors may be used simultaneously.

[0027] As described above, in the magnetization device degradation diagnosis system 200 of Embodiment 2, when the sensor 3 is fixed to the radial outer surface of the magnetization yoke 1, the sensor 3 is fixed at a position where the distance L1 from the magnetization coil 5 to the sensor 3 is shortest. Therefore, it is possible to suppress the attenuation of the signal generated when current is supplied to the magnetization coil 5 before it is detected by the sensor 3.

[0028] As described above, the magnetization device degradation diagnosis system of Embodiment 2 can detect failures in the magnetization device in advance without affecting the energizing conditions to the magnetization yoke, thereby preventing a decrease in productivity due to sudden failures of the magnetization device. Furthermore, it can suppress the attenuation of signals generated during energization before they are detected by the sensor.

[0029] Embodiment 3. Embodiment 3 involves fixing the sensor to the outer axial surface of the magnetization yoke at a position where the distance from the magnetization coil is minimized.

[0030] The following describes the degradation diagnosis system for the magnetization device of Embodiment 3, focusing on the differences from Embodiment 1, based on Figure 4. Figures 1 and 2 of Embodiment 1 will be referenced as appropriate. In the drawings of Embodiment 3, parts that are the same as or corresponding to those in Embodiment 1 are denoted by the same reference numerals. To distinguish it from Embodiment 1, the magnetization device 30 and the magnetization device deterioration diagnosis system 300 are used.

[0031] Figure 4 is a longitudinal cross-sectional view showing the structure of the magnetization yoke 1 in the magnetization device deterioration diagnosis system 300 according to Embodiment 3. The magnetization device deterioration diagnosis system 300 of Embodiment 3 includes a magnetization yoke 1 in which a magnetization coil 5 is wound around an iron core 4, a magnetization power supply 2 connected to the magnetization yoke 1, a sensor 3 fixed to the magnetization yoke 1, and a signal processing device 6 connected to the sensor 3.

[0032] A feature of the magnetization device deterioration diagnosis system 300 of this third embodiment is that, when the sensor 3 is fixed to the outer peripheral end face in the axial direction of the magnetization yoke 1, the sensor 3 is fixed at a position where the distance L2 from the magnetization coil 5 to the sensor 3 is shortest.

[0033] Regarding the position of the magnetizing coil 5, when the distance to the axial end face is smaller than the distance to the radial outer surface, axial installation can suppress the distance between the magnetizing coil 5 and the sensor, allowing for the acquisition of good data. In particular, if the magnetizing coil 5 has a cooling structure (using water and coolant) on its radially outer side, axial installation allows for the acquisition of more sensitive and better data.

[0034] The type of sensor 3 used is not limited to those that directly detect sound or vibration. For example, a configuration using an AE sensor that detects elastic waves, or an acceleration sensor or strain gauge, may be used as a sensor to detect deformation and cracks. Furthermore, the number of sensors is not limited to the single sensor shown in Figure 4. For example, multiple sensors may be fixed together in one location, or one sensor may be fixed in multiple locations. When using multiple sensors, multiple types of sensors may be used simultaneously.

[0035] As described above, in the magnetization device deterioration diagnosis system 300 of Embodiment 3, when the sensor 3 is fixed to the outer axial surface of the magnetization yoke 1, the sensor 3 is fixed at a position where the distance L2 from the magnetization coil 5 to the sensor 3 is shortest. Therefore, it is possible to suppress the attenuation of the signal generated when current is supplied to the magnetization coil 5 before it is detected by the sensor 3.

[0036] As described above, the magnetization device degradation diagnosis system of Embodiment 3 can detect failures in the magnetization device in advance without affecting the energizing conditions to the magnetization yoke, thereby preventing a decrease in productivity due to sudden failures of the magnetization device. Furthermore, it can suppress the attenuation of signals generated during energization before they are detected by the sensor.

[0037] Embodiment 4. Embodiment 4 involves fixing the sensor to the magnetization yoke using a sensor fixing jig.

[0038] The following describes the degradation diagnosis system for the magnetization device of Embodiment 4, focusing on the differences from Embodiment 1, based on the respective figures. Figures 1 and 2 of Embodiment 1 will be referenced as appropriate. In the drawings of Embodiment 4, parts that are the same as or corresponding to those in Embodiment 1 are denoted by the same reference numerals. For the purpose of distinguishing this from Embodiment 1, the magnetization device 40 and the magnetization device deterioration diagnosis system 400 are used.

[0039] Figure 5 is a schematic diagram showing the configuration of the magnetization device deterioration diagnosis system 400 according to Embodiment 4. Figure 6A is an explanatory diagram of the sensor fixing jig 11 for installing the sensor 3 on the magnetization yoke 1. Figure 6B is an explanatory diagram for installing the sensor fixing jig 11 for installing the sensor 3 on the magnetization yoke 1. The magnetization device deterioration diagnosis system 400 of Embodiment 4 includes a magnetization yoke 1 in which a magnetization coil 5 is wound around an iron core 4, a magnetization power supply 2 connected to the magnetization yoke 1, a sensor 3 fixed to the magnetization yoke 1 via a sensor fixing jig 11, and a signal processing device 6 connected to the sensor 3.

[0040] A feature of the magnetization device deterioration diagnosis system 400 of this embodiment 4 is that the sensor 3 is fixed to the magnetization yoke 1 via a sensor fixing jig 11 that surrounds substantially the entire circumference of the magnetization yoke 1 in the circumferential direction. The position in which the sensor fixing jig 11 is attached to the magnetization yoke 1 is not limited to the vicinity of the axial center as shown in Figure 5. For example, it may be attached in a position where the axial end face and the end face of the sensor fixing jig 11 are aligned. Furthermore, the number of sensor fixing jigs 11 and the number of sensors 3 fixed to each sensor fixing jig 11 are not limited to one as shown in Figure 5. For example, multiple sensor fixing jigs 11 may be attached to the magnetization yoke 1, and multiple sensors 3 may be fixed to each sensor fixing jig 11. When using multiple sensors, multiple types of sensors 3 may be used simultaneously.

[0041] The sensor fixing jig 11, which secures the sensor 3 to the magnetizing yoke 1, is in contact with the outer surface of the magnetizing yoke 1 at its inner circumferential surface. Note that the configuration is not limited to the one shown in Figure 5, where the inner surface of the sensor fixing jig 11 is in full contact with the outer surface of the magnetization yoke 1. For example, the configuration may be one in which only a part of the inner surface of the sensor fixing jig 11 and the outer surface of the magnetization yoke 1 are in contact.

[0042] The shape of the sensor fixing jig 11 is not limited to the annular shape shown in Figure 5. For example, if the magnetization yoke 1 has a polygonal shape, the jig may have a polygonal shape that matches it. Furthermore, the component configuration of the sensor fixing jig 11 is not limited to the divided shape shown in Figures 6A and 6B. For example, it may be an annular shape without divisions or a shape divided in the axial direction. Furthermore, the number of divisions of the sensor fixing jig 11 is not limited to the two shown in Figure 6. Also, the division positions when dividing into two are not limited to the positions shown in Figures 6A and 6B. For example, it may be divided into two arc shapes with central angles of 240° and 120°.

[0043] In the configuration of the sensor fixing jig 11 shown in Figures 6A and 6B, the divided sensor fixing jigs 11 are in contact with each other over a surface area, but the configuration is not limited to the sensor fixing jigs 11 being in contact with each other. When the sensor fixing jig 11 is composed of multiple parts, the sensor fixing jig 11 and the magnetization yoke 1 are in contact, but there may be a gap between the sensor fixing jigs 11 and they are not in contact with each other. Furthermore, even if the sensor fixing jig 11 is composed of a single ring-shaped component, it may be given a shape with a slit in part.

[0044] As described above, in the magnetization device degradation diagnosis system 400 of Embodiment 4, the sensor 3 is fixed via a sensor fixing jig 11 that surrounds substantially the entire circumference of the magnetization yoke 1. Therefore, it is possible to detect signals that have propagated through the sensor fixing jig 11, not just signals generated from the magnetization coil 5 near the sensor 3. Furthermore, by shaping the sensor fixing jig 11 to match the shape of the magnetization yoke 1, or to avoid interference with surrounding equipment, the sensor 3 can be fixed regardless of the shape of the existing magnetization yoke and equipment configuration. Therefore, it can be applied not only when introducing new magnetization equipment but also to existing magnetization devices for deterioration diagnosis.

[0045] As described above, the magnetization device degradation diagnosis system of Embodiment 4 can detect failures in the magnetization device in advance without affecting the energizing conditions to the magnetization yoke, thereby preventing a decrease in productivity due to sudden failures of the magnetization device. Furthermore, it is not limited to signals generated from the magnetization coil near the sensor, but can also detect signals propagated through the sensor fixing jig. In addition, it can be applied to existing magnetization devices to perform degradation diagnosis.

[0046] Embodiment 5. Embodiment 5 determines the deterioration of the magnetization device by comparing the sensor output recorded each time power is applied with a set threshold. This threshold is set using machine learning.

[0047] The following describes the degradation diagnosis system for the magnetization device of Embodiment 5, focusing on the differences from Embodiment 1, based on the respective figures. Note that the configuration of the degradation diagnosis system for the magnetization device is the same as in Embodiment 1, so please refer to Figures 1 and 2 of Embodiment 1 as appropriate.

[0048] Figure 7 shows an example of sensor output in the magnetization device degradation diagnosis system according to Embodiment 5. The horizontal axis represents the passage of time, and the vertical axis represents the magnitude of the sensor output. The scale (G) on the horizontal axis indicates the timing of energization to the magnetization yoke 1. Figure 7 shows an example of the sensor output waveform SGW and a pre-set threshold THV when energization is applied at regular time intervals.

[0049] In the magnetization device degradation diagnosis system of Embodiment 5, the output of the sensor 3 is recorded each time current is supplied to the magnetization yoke 1 by the sensor 3 fixed to the magnetization yoke 1 and the signal processing device 6. Furthermore, each time the sensor output is recorded, its magnitude is compared to a preset threshold THV.

[0050] A feature of the magnetization device degradation diagnosis system of Embodiment 5 is that when the sensor output recorded each time current is supplied to the magnetization yoke 1 exceeds a preset threshold THV, it is determined that the magnetization yoke 1 has deteriorated and that replacement of the magnetization yoke 1 is necessary. Regarding the determination of magnetization yoke 1 degradation by comparing sensor output and threshold THV, the timing for determining that replacement is necessary is not limited to when the sensor output first exceeds the threshold THV. For example, the judgment criterion could be defined as the sensor output exceeding the threshold THV a predetermined number of times consecutively.

[0051] As described above, the magnetization device deterioration diagnosis system of Embodiment 5 determines the deterioration of the magnetization yoke 1 by comparing the sensor output recorded each time power is applied with a preset threshold THV, and determines whether replacement is necessary. This allows for the arrangement of a replacement magnetizing yoke 1 at an appropriate time, before a failure of the magnetizing yoke 1 occurs. Therefore, not only is productivity reduced due to equipment downtime, but the accumulation of excess inventory can be prevented by arranging for replacements early.

[0052] Here, we will explain how to set the threshold THV when using machine learning. Figure 8 is a diagram showing the configuration of a learning device 51 related to a degradation diagnosis system for a magnetization device. The learning device 51 comprises a data acquisition unit 52, a model generation unit 53, and a learned model storage unit 54.

[0053] The data acquisition unit 52 acquires information on the energization of the magnetization yoke 1 and the sensor output obtained during energization as training data. The model generation unit 53 learns the sensor output during energization based on training data created from a combination of energization information for the magnetization yoke 1 output from the data acquisition unit 52 and the sensor output obtained during energization. In other words, a trained model is generated that infers the sensor output during energization from the energization information to the magnetization yoke 1 of the magnetization device degradation diagnosis system and the sensor output obtained during energization. Here, the training data is data that correlates the energization information to the magnetization yoke 1 and the sensor output obtained during energization.

[0054] As information about the energization of the magnetizing yoke 1, for example, the number of times the magnetizing yoke 1 has been energized can be used. As shown in Figure 7, the amplitude of the sensor output of the magnetization yoke 1 is small in its initial state and tends to gradually increase with use. The higher the voltage applied to the magnetization coil 5 when energized, the more prone it is to deterioration, and the higher the current applied when energized, the more prone it is to deterioration. For this reason, the voltage applied to the magnetization coil 5 and the current applied when energized can be used as energization information.

[0055] The model generation unit 53 can use any known learning algorithm, such as supervised learning, unsupervised learning, or reinforcement learning. As an example, the case where a neural network is applied will be described.

[0056] The model generation unit 53 learns the sensor output when power is applied, for example, by supervised learning according to a neural network model. Here, supervised learning refers to a method in which a learning device is given pairs of input and result (label) data, learns features in that training data, and infers the result from the input.

[0057] A neural network consists of an input layer made up of multiple neurons, an intermediate layer (hidden layer) made up of multiple neurons, and an output layer made up of multiple neurons. The intermediate layer can be one or more layers. For example, in a three-layer neural network like the one shown in Figure 9, when multiple inputs are input to the input layer (X1-X3), these values ​​are multiplied by weights W1 (w11-w16) and input to the hidden layer (Y1-Y2). The result is then further multiplied by weights W2 (w21-w26) and output from the output layer (Z1-Z3). This output varies depending on the values ​​of weights W1 and W2.

[0058] In this embodiment 5, the neural network learns the sensor output during energization by so-called supervised learning, according to training data created based on the combination of energization information for the magnetization yoke 1 acquired by the data acquisition unit 52 and the sensor output obtained during energization. In other words, the neural network learns by inputting information about the energization of the magnetized yoke 1 into the input layer and adjusting the weights W1 and W2 so that the result output from the output layer approaches the sensor output obtained when the energy is applied. The model generation unit 53 generates and outputs a trained model by performing the learning described above. The trained model storage unit 54 stores the trained model output from the model generation unit 53.

[0059] Next, the learning process of the learning device 51 will be explained based on Figure 10. Figure 10 is a flowchart of the learning process of the learning device 51.

[0060] In step S11, the data acquisition unit 52 acquires information on the energization of the magnetization yoke 1 and the sensor output obtained when energization is performed. Although the energization information for the magnetization yoke 1 and the sensor output obtained when energization is performed are acquired simultaneously, it is sufficient if the energization information for the magnetization yoke 1 and the sensor output obtained when energization is performed are input in association, and the data for the energization information for the magnetization yoke and the sensor output obtained when energization is performed may be acquired at different times. In step S12, the model generation unit 53 learns the sensor output during energization through so-called supervised learning, according to training data created based on the combination of energization information to the magnetization yoke 1 acquired by the data acquisition unit 52 and the sensor output obtained during energization, and generates a trained model. In step S13, the trained model storage unit 54 stores the trained model generated by the model generation unit 53.

[0061] Next, we will describe the inference device 55 that uses a trained model to infer the sensor output during power application from the power application information. Figure 11 is a diagram showing the configuration of an inference device 55 related to a degradation diagnosis system for a magnetization device. The inference device 55 comprises a data acquisition unit 56, an inference unit 57, and a trained model storage unit 58. The data acquisition unit 56 acquires information on the energization of the magnetization yoke. The inference unit 57 infers the sensor output during energization, obtained using a trained model. That is, by inputting the energization information to the magnetization yoke 1 acquired by the data acquisition unit 56 into this trained model, it is possible to output the sensor output during energization inferred from the energization information to the magnetization yoke 1. In this embodiment 5, the sensor output during energization was described as being output using a trained model learned by the model generation unit 53 of the magnetization device degradation diagnosis system. However, a trained model may be obtained from another magnetization device degradation diagnosis system, and the sensor output during energization may be output based on this trained model.

[0062] Next, based on Figure 12, we will explain the process for obtaining the sensor output when power is applied using the inference device 55. In step S21, the data acquisition unit 56 acquires information on the energization of the magnetization yoke 1. In step S22, the inference unit 57 retrieves a trained model from the trained model storage unit 58, inputs the energization information to the magnetization yoke 1 to the retrieved trained model, and obtains the sensor output when energized. In step S23, the inference unit 57 outputs the sensor output during energization, obtained by the trained model, to the magnetization device degradation diagnosis system.

[0063] The threshold THV can be determined based on the sensor output during energization obtained by the inference unit 57. For example, regarding the sensor output obtained by the inference unit 57 when power is applied, when the output changes from its initial state, the value that was not exceeded before the change but is exceeded after the change can be determined as a threshold for determining whether or not a change has occurred.

[0064] In this embodiment 5, we have described the case in which supervised learning is applied to the learning algorithm used by the model generation unit 53. However, this is not the only possible case. Regarding learning algorithms, in addition to supervised learning, reinforcement learning, unsupervised learning, or semi-supervised learning can also be applied. Furthermore, the model generation unit may learn the sensor output during energization according to training data created for degradation diagnostic systems of multiple magnetization devices. The model generation unit may also acquire training data from multiple degradation diagnostic systems used in the same region. Alternatively, it may learn the sensor output during energization by utilizing training data collected from multiple degradation diagnostic systems of magnetization devices operating independently in different locations. Furthermore, it is possible to add or remove the degradation diagnosis system for the magnetization device that collects training data from the target during the process. Furthermore, a learning device 51 that has learned the sensor output during energization for a degradation diagnosis system of a certain magnetizing device may be applied to a degradation diagnosis system of a different magnetizing device, and the sensor output during energization for this other magnetizing device degradation diagnosis system may be relearned and updated.

[0065] Furthermore, the learning algorithm used in the model generation unit 53 can be deep learning, which learns to extract the features themselves, or machine learning can be performed according to other known methods, such as genetic programming, inductive logic programming, or support vector machines. The output of sensor 3 may be recorded in a database along with usage records each time current is applied to the magnetization yoke 1. Machine learning may be performed based on the information in the database to generate a model for obtaining the threshold THV.

[0066] Furthermore, the learning device 51 and the inference device 55 are intended to be used as part of a degradation diagnosis system for the magnetization device to learn the sensor output when power is applied. For example, the learning device 51 and the inference device 55 may be built into the signal processing device 6. However, the magnetization device may be connected to a degradation diagnosis system via a network, and may be a separate device from this degradation diagnosis system. Furthermore, the learning device 51 and the inference device 55 may reside on a cloud server.

[0067] As described above, the magnetization device degradation diagnosis system of Embodiment 5 can detect failures in the magnetization device in advance without affecting the energizing conditions to the magnetization yoke, thereby preventing a decrease in productivity due to sudden failures of the magnetization device. Furthermore, it can detect abnormalities in the magnetization yoke at an appropriate time before a failure occurs and take appropriate action. In addition, it can set more objective criteria for determining the degradation of the magnetization yoke.

[0068] Embodiment 6. Embodiment 6 diagnoses the deterioration of the magnetization yoke based on changes in the shape of the sensor output waveform.

[0069] The following describes the degradation diagnosis system for the magnetization device of Embodiment 6, focusing on the differences from Embodiment 1, based on the respective figures. Note that the configuration of the degradation diagnosis system for the magnetization device is the same as in Embodiment 1, so please refer to Figures 1 and 2 of Embodiment 1 as appropriate.

[0070] The magnetization device degradation diagnosis system of Embodiment 6 comprises a magnetization yoke 1 in which a magnetization coil 5 is wound around an iron core 4, a magnetization power supply 2 connected to the magnetization yoke 1, a sensor 3 installed on the magnetization yoke 1 and normally fixed, and a signal processing device 6 connected to the sensor 3.

[0071] Figure 13 shows an example of sensor output in the magnetization device degradation diagnosis system according to Embodiment 6. Figure 14 shows another example of sensor output in the magnetization device degradation diagnosis system according to Embodiment 6. In both Figures 13 and 14, the horizontal axis represents the passage of time, and the vertical axis represents the magnitude of the sensor output. The scale (G) on the horizontal axis indicates the timing of energization to the magnetization yoke 1. Figures 13 and 14 also show examples of the sensor output waveform SGW when energization is repeated at regular time intervals.

[0072] In the magnetization device degradation diagnosis system of Embodiment 6, the sensor 3 fixed to the magnetization yoke 1 and the signal processing device 6 record the change in the sensor output over time while the magnetization yoke 1 is repeatedly energized. Then, the presence or absence of a change in the time-series data is determined from the recorded sensor output. Regarding changes in time-series data, the criterion for determination may be whether the magnitude of the sensor output changes over time, as shown in Figure 13. In this case, the change is not limited to changes in the direction of increasing magnitude of the sensor output, as shown in Figure 13. For example, a decrease in the magnitude of the sensor output may also be considered a change.

[0073] On the other hand, as shown in Figure 14, the change in the shape of the sensor output waveform SGW over time may be used as the criterion for determination. In this case, the change is not limited to an increase in the number of peaks in the sensor output waveform SGW as shown in Figure 14. For example, in the sensor output waveform SGW, the change may be determined by the difference in the time required for the magnitude of the sensor output to return to a normal value after reaching a peak value. Furthermore, both the magnitude of the sensor output and the shape of the sensor output waveform SGW may be used as criteria for determining changes in the sensor output waveform SGW. Furthermore, machine learning as described in Embodiment 5 may be used for processing time-series data in Embodiment 6.

[0074] The characteristic of the magnetization device deterioration diagnosis system of this embodiment 6 is that it determines the deterioration of the magnetization yoke 1 from the changes in time-series data of the sensor output recorded while energizing the magnetization yoke 1 repeatedly, and determines that replacement of the magnetization yoke 1 is necessary.

[0075] As described above, the magnetization device degradation diagnosis system in Embodiment 6 determines whether the magnetization yoke 1 needs to be replaced based on changes in time-series data of sensor output recorded while the magnetization yoke 1 is repeatedly energized. This allows for the procurement of a replacement magnetization yoke 1 at an appropriate time before a failure occurs. Therefore, it is possible to prevent not only a decrease in productivity due to equipment downtime, but also the occurrence of excess inventory by procuring replacement parts early.

[0076] As described above, the magnetization device degradation diagnosis system of Embodiment 6 can detect failures in the magnetization device in advance without affecting the energizing conditions to the magnetization yoke, thereby preventing a decrease in productivity due to sudden failures of the magnetization device. Furthermore, it can detect and address abnormalities in the magnetization yoke at an appropriate time before a failure of the magnetization yoke occurs.

[0077] Embodiment 7. Embodiment 7 predicts the lifespan of the magnetization yoke each time it is energized.

[0078] The following describes the degradation diagnosis system for the magnetization device of Embodiment 7, focusing on the differences from Embodiment 1, based on Figure 15. Since the configuration of the magnetization device degradation diagnosis system is the same as in Embodiment 1, please refer to Figures 1 and 2 of Embodiment 1 as appropriate.

[0079] Figure 15 is a flowchart showing the operation of the magnetization device degradation diagnosis system according to Embodiment 7. The magnetization device degradation diagnosis system of Embodiment 7 comprises a magnetization yoke 1 in which a magnetization coil 5 is wound around an iron core 4, a magnetization power supply 2 connected to the magnetization yoke 1, a sensor 3 installed on the magnetization yoke 1 and normally fixed, and a signal processing device 6 connected to the sensor 3.

[0080] The characteristic of the magnetization device degradation diagnosis system of this embodiment 7 is that it takes the sensor output recorded each time current is supplied to the magnetization yoke 1 as input and predicts the lifespan until the magnetization device fails. This makes it possible to predict in advance when the magnetization yoke 1 may fail, preventing equipment shutdowns and decreased productivity caused by sudden failures of the magnetization yoke 1. Furthermore, by recording the change in estimated lifespan for each operating condition of the magnetization device, it is possible to determine the operating conditions that extend the lifespan of the magnetization yoke 1.

[0081] The operation of the degradation diagnosis system for the magnetization device in Embodiment 7 will be explained based on the flowchart in Figure 15. In step S31, it is first determined whether or not magnetization has been performed. If executed, proceed to step S32; otherwise, terminate the process. In step S32, the magnetization coil 5 is energized. In step S33, the sensor output waveform SGW is acquired. In step S34, the estimated lifespan of the magnetization yoke 1 is calculated based on the acquired sensor output waveform SGW. In step S35, the estimated lifespan of the magnetized yoke 1 is output. In step S36, the calculated estimated lifespan is compared with a pre-set threshold. If the estimated lifespan is less than the threshold, i.e., if the lifespan is insufficient, a decision is made to discontinue use, and the process proceeds to step S37. On the other hand, if the estimated lifetime is above the threshold, the process returns to step S31 and the above process is repeated. In step S37, the use of the magnetization device 10 is stopped.

[0082] Furthermore, the method for calculating the estimated lifespan based on the degradation diagnosis system sensor output of the magnetization device in Embodiment 7 is not limited to a specific method. For example, the lifespan may be estimated by comparing the usage period until failure with the changes in sensor output up to that point in the usage history of magnetization yokes 1 of the same shape or a similar structure, and determining which period from the start of use to failure the trend in sensor output of the magnetization yoke 1 in use corresponds to. Furthermore, machine learning may be used to process the sensor output data in Embodiment 7. For example, data on the usage history of the magnetization yoke 1 (time until failure) and the sensor output at the time of failure (immediately before failure) can be collected, and this data can be used to train a model of the relationship between the lifespan of the magnetization yoke and the sensor output.

[0083] As described above, the magnetization device degradation diagnosis system of Embodiment 7 can detect failures in the magnetization device in advance without affecting the energizing conditions to the magnetization yoke, thereby preventing a decrease in productivity due to sudden failures of the magnetization device. Furthermore, it is possible to predict the lifespan of the magnetization yoke each time it is energized and replace the magnetization yoke at an appropriate time before a failure occurs.

[0084] Embodiment 8. Embodiment 8 enables the detection of initial defects during magnetization yoke replacement.

[0085] The following describes the degradation diagnosis system for the magnetization device of Embodiment 8, focusing on the differences from Embodiment 1. Note that the configuration of the degradation diagnosis system for the magnetization device is the same as in Embodiment 1, so please refer to Figures 1 and 2 of Embodiment 1 as appropriate.

[0086] The magnetization device degradation diagnosis system of Embodiment 8 comprises a magnetization yoke 1 in which a magnetization coil 5 is wound around an iron core 4, a magnetization power supply 2 connected to the magnetization yoke 1, a sensor 3 installed on the magnetization yoke 1 and normally fixed, and a signal processing device 6 connected to the sensor 3.

[0087] A key feature of the magnetization device degradation diagnosis system in this embodiment 8 is that it detects initial defects before use by recording and comparing the sensor output during the first power-on when the magnetization yoke 1 is newly installed and immediately after replacement. Specifically, when the magnetization yoke 1 is replaced due to a malfunction or other reason, the presence or absence of an initial defect is determined by comparing it with the sensor output recorded up to that point. This allows for the determination of whether or not there are defects in the replacement part, thus preventing sudden failures due to initial defects before use. For determining whether a replacement part is faulty, the system may use criteria based on either the magnitude of the sensor output, the shape of the sensor output waveform (SGW), or both.

[0088] Furthermore, since the sensor output is used to determine initial defects, power must be supplied. However, the current value can be set as the test value, which may differ from the actual operating conditions of the magnetized yoke 1. Specifically, for example, to prevent damage to the magnetized yoke 1 during testing, a current smaller than the actual operating conditions can be set as the test value.

[0089] As described above, the magnetization device degradation diagnosis system of Embodiment 8 can detect failures in the magnetization device in advance without affecting the energizing conditions to the magnetization yoke, thereby preventing a decrease in productivity due to sudden failures of the magnetization device. Furthermore, it can detect initial defects when replacing the magnetization yoke, thus avoiding sudden failures.

[0090] Embodiment 9. Embodiment 9 focuses on the magnetizing device of the degradation diagnosis system for the magnetizing device described in Embodiment 1.

[0091] The magnetization device of Embodiment 9 will be described below with reference to Figures 1 and 2 of Embodiment 1, Figure 3 of Embodiment 2, Figure 4 of Embodiment 3, and Figure 5 of Embodiment 4. The magnetization device 10 of Embodiment 9 includes a magnetization yoke 1 in which a magnetization coil 5 is wound around an iron core 4, a magnetization power supply 2 connected to the magnetization yoke 1, a sensor 3 installed on the magnetization yoke 1 and normally fixed, and a signal processing device 6 connected to the sensor 3.

[0092] A feature of the magnetization device 10 of Embodiment 9 is that it has a sensor 3 for detecting signals caused by sound or vibration generated when current is supplied to the magnetization coil 5 of the magnetization yoke 1. By having the magnetization device 10 equipped with this sensor 3, a deterioration diagnosis system can be constructed. By recording the sensor output each time current is supplied to the magnetization coil 5, it is possible to determine the deterioration of the magnetization yoke and predict its lifespan.

[0093] The type of sensor 3 is not limited to those that directly detect sound or vibration. For example, a configuration using an AE sensor that detects elastic waves, or an acceleration sensor or strain gauge, may be used as a sensor to detect deformation and cracks.

[0094] Furthermore, the number of sensors 3 is not limited to one. For example, multiple sensors may be fixed together in one location, or one sensor may be fixed in multiple locations. When using multiple sensors, multiple types of sensors may be used simultaneously. Regarding the fixing position of the sensor 3, it can be fixed to the radial outer surface of the magnetization yoke 1 as described in Embodiment 2, or to the axial outer surface of the magnetization yoke 1 as described in Embodiment 3. Furthermore, the sensor 3 can be installed using the sensor fixing jig 11 described in Embodiment 4. Additionally, the sensor 3 may be embedded inside the magnetization yoke 1.

[0095] As described above, the magnetization device of Embodiment 9 can detect a malfunction of the magnetization device in advance without affecting the energization conditions to the magnetization yoke, and can prevent a decrease in productivity due to a sudden failure of the magnetization device.

[0096] Here, an example of the hardware of the signal processing device 6, learning device 51, and inference device 55 in embodiments 1, 4, and 5 will be described with reference to Figure 16, using the signal processing device 6 as a representative example. The signal processing device 6 consists of a processor 1000 and a storage device 1001. The storage device 1001 includes a volatile storage device such as a random access memory (not shown) and a non-volatile auxiliary storage device such as a flash memory. Alternatively, a hard disk may be provided as an auxiliary storage device instead of flash memory. The processor 1000 executes a program input from the storage device 1001. In this case, the program is input to the processor 1000 from the auxiliary storage device via a volatile storage device. The processor 1000 may also output data such as calculation results to the volatile storage device of the storage device 1001, or it may save the data to the auxiliary storage device via the volatile storage device.

[0097] While this disclosure describes various exemplary embodiments and examples, the various features, aspects, and functions described in one or more embodiments are not limited to the application of a particular embodiment, but are applicable to the embodiments individually or in various combinations. Accordingly, countless variations not illustrated are conceivable within the scope of the art disclosed in this specification. These include, for example, modifying, adding or omitting at least one component, or even extracting at least one component and combining it with components of other embodiments.

[0098] The various aspects of this disclosure are summarized below as an appendix.

[0099] (Note 1) The device comprises a magnetizing yoke in which a magnetizing coil is wound around an iron core, a magnetizing power supply connected to the magnetizing coil, a sensor installed on the magnetizing yoke, and a signal processing device connected to the sensor. A deterioration diagnosis system for a magnetizing device, wherein the sensor detects a signal caused by sound or vibration generated when current is supplied from the magnetizing power supply to the magnetizing coil, and the signal processing device diagnoses the deterioration state of the magnetizing yoke based on the output of the sensor. (Note 2) The deterioration diagnosis system for the magnetization device described in Appendix 1, wherein the sensor is installed at a position that is the shortest distance from the magnetization coil. (Note 3) The deterioration diagnosis system for a magnetizing device as described in Appendix 1 or Appendix 2, wherein the sensor is installed on the magnetizing yoke via a separate component that surrounds the outer circumferential surface of the magnetizing yoke. (Note 4) The signal processing device compares the output value of the sensor with a set threshold value to determine the timing for replacing the magnetization yoke. This is a deterioration diagnosis system for a magnetization device according to any one of the appendices 1 to 3. (Note 5) The signal processing device is a deterioration diagnosis system for a magnetizing device according to any one of the appendices 1 to 3, which determines the replacement timing of the magnetizing yoke from changes in the time-series data of the output of the sensor. (Note 6) The signal processing device is a deterioration diagnostic system for a magnetizing device according to any one of the appendices 1 to 3, which takes the output of the sensor recorded each time the magnetizing yoke is energized as input to predict the lifespan of the magnetizing yoke. (Note 7) A data acquisition unit that acquires learning data including information on the energization of the magnetization yoke and the output of the sensor obtained when energization is performed, A model generation unit generates a trained model for inferring the output of the sensor during energization from the energization information of the magnetization yoke using the aforementioned training data, A deterioration diagnosis system for a magnetizing device as described in any one of the appendices 1 to 3, comprising the features described herein. (Note 8) A data acquisition unit that acquires information on the energization of the magnetization yoke, An inference unit that uses a trained model to infer the output of the sensor when energized from the energization information of the magnetization yoke, and outputs a predicted sensor output when energized from the energization information of the magnetization yoke acquired by the data acquisition unit, A deterioration diagnosis system for a magnetizing device as described in any one of the appendices 1 to 3, comprising the features described herein. (Note 9) The signal processing device compares the expected sensor output during energization with the output of the sensor during energization to determine the deterioration of the magnetization yoke, as described in Appendix 8. (Note 10) The signal processing device is The output of the sensor during energization when the magnetization yoke is newly installed is recorded. The output of the sensor during the first energization immediately after the replacement of the magnetization yoke is recorded. A magnetization device deterioration diagnosis system according to any one of the appendices 1 to 3, which, when replacing the magnetization yoke, compares the output of the sensor when energized upon new installation of the magnetization yoke with the output of the sensor when energized immediately after replacement to determine whether the magnetization yoke is good or bad. (Note 11) A magnetizing device comprising a magnetizing yoke in which a magnetizing coil is wound around an iron core, a magnetizing power supply connected to the magnetizing coil, a sensor installed on the magnetizing yoke for detecting signals caused by sound or vibration generated when current is supplied from the magnetizing power supply to the magnetizing coil, and a signal processing device connected to the sensor. (Note 12) The magnetization device described in Appendix 11 is installed in a position that minimizes the distance between the sensor and the magnetization coil. (Note 13) The magnetization device according to Appendix 11 or Appendix 12, wherein the sensor is installed on the magnetization yoke via a separate component that surrounds the outer circumferential surface of the magnetization yoke. [Explanation of Symbols]

[0100] 1. Magnetizing yoke, 2. Magnetizing power supply, 3. Sensor, 4. Iron core, 5. Magnetizing coil, 6 Signal processing device, 10, 20, 30, 40 Magnetizing device, 11 Sensor fixing jig, 51 Learning device, 52 Data acquisition unit, 53 Model generation unit, 54 Pre-trained model storage unit, 55 Inference unit, 56 Data acquisition unit, 57 Inference unit, 58. Trained model memory unit, 100, 200, 300, 400 Magnetization Device Degradation Diagnostic System 1000 processors, 1001 memory devices.

Claims

1. The device comprises a magnetizing yoke in which a magnetizing coil is wound around an iron core, a magnetizing power supply connected to the magnetizing coil, a sensor installed on the magnetizing yoke, and a signal processing device connected to the sensor. A deterioration diagnosis system for a magnetizing device, wherein the sensor detects a signal caused by sound or vibration generated when current is supplied from the magnetizing power supply to the magnetizing coil, and the signal processing device diagnoses the deterioration state of the magnetizing yoke based on the output of the sensor.

2. The deterioration diagnosis system for a magnetizing device according to claim 1, wherein the sensor is installed at a position that is closest to the magnetizing coil.

3. A deterioration diagnosis system for a magnetizing device according to claim 1 or claim 2, wherein the sensor is installed on the magnetizing yoke via a separate component that surrounds the outer surface of the magnetizing yoke.

4. The signal processing device compares the output value of the sensor with a set threshold value to determine the timing for replacing the magnetization yoke. This is a deterioration diagnosis system for a magnetization device according to claim 1 or claim 2.

5. The signal processing device determines the timing for replacing the magnetization yoke from the change in time-series data of the output of the sensor, in the magnetization device deterioration diagnosis system according to claim 1 or claim 2.

6. The signal processing device predicts the lifespan of the magnetizing yoke using the output of the sensor recorded each time the magnetizing yoke is energized as input, in the deterioration diagnosis system for a magnetizing device according to claim 1 or 2.

7. A data acquisition unit that acquires learning data including information on the energization of the magnetization yoke and the output of the sensor obtained when energization is performed, A model generation unit generates a trained model for inferring the output of the sensor during energization from the energization information of the magnetization yoke using the aforementioned training data, A deterioration diagnosis system for a magnetizing device according to claim 1 or claim 2, comprising:

8. A data acquisition unit that acquires information on the energization of the magnetization yoke, An inference unit that uses a trained model to infer the output of the sensor when energized from the energization information of the magnetization yoke, and outputs a predicted sensor output when energized from the energization information of the magnetization yoke acquired by the data acquisition unit, A deterioration diagnosis system for a magnetizing device according to claim 1 or claim 2, comprising:

9. The signal processing device compares the expected sensor output during energization with the output of the sensor during energization to determine the deterioration of the magnetization yoke, as described in claim 8.

10. The signal processing device is The output of the sensor during energization when the magnetization yoke is newly installed is recorded. The output of the sensor during the first energization immediately after the replacement of the magnetization yoke is recorded. A deterioration diagnosis system for a magnetizing device according to claim 1 or 2, which, when replacing the magnetizing yoke, compares the output of the sensor when energized upon new installation of the magnetizing yoke with the output of the sensor when energized immediately after replacement to determine whether the magnetizing yoke is good or bad.

11. A magnetizing device comprising a magnetizing yoke in which a magnetizing coil is wound around an iron core, a magnetizing power supply connected to the magnetizing coil, a sensor installed on the magnetizing yoke for detecting signals caused by sound or vibration generated when current is supplied from the magnetizing power supply to the magnetizing coil, and a signal processing device connected to the sensor.

12. The magnetization device according to claim 11, wherein the sensor is installed at a position that minimizes the distance from the magnetization coil.

13. The magnetization device according to claim 11 or claim 12, wherein the sensor is installed on the magnetization yoke via a separate component that surrounds the outer circumferential surface of the magnetization yoke.

Images