Detection device, abnormality detection system, and abnormality detection method

Abstract

To detect a state of an operation environment of a power module with a simpler configuration.SOLUTION: An abnormality detection device includes a loss calculation unit, a time differential value calculation unit, and an identification unit. The loss calculation unit calculates a loss pattern of the power module over time. The time differential value calculation unit calculates a time differential value of the temperature at the measurement position. The identification unit identifies an abnormality corresponding to a first threshold by comparing the first threshold determined according to the loss pattern in a first time period with a first extreme value of an absolute value of a time differential value of the temperature in the first time period.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] FIELD Embodiments of the present invention relate to a detection device, an anomaly detection system, and an anomaly detection method. [Background technology]

[0002] It has long been known that heat is one of the main causes of failure in power electronics equipment. The higher the temperature, the faster parts tend to break down. Furthermore, even when the equipment is operating at temperatures below the specified value, the higher the temperature, the faster parts tend to break down. Due to design constraints and other reasons such as limitations on sensor placement, it is rare for sensors to be able to directly detect thermal conditions (such as the temperature of critical parts and cooling performance). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 6615329 [Patent Document 2] Patent No. 3668708 Summary of the Invention [Problem to be solved by the invention]

[0004] Conventional techniques have not been able to detect the state of the operating environment of a power module with a simpler configuration. [Means for solving the problem]

[0005] A detection device according to an embodiment includes a loss calculation unit, a time derivative calculation unit, and an identification unit. The loss calculation unit calculates a loss pattern of a power module over time. The time derivative calculation unit calculates a time derivative of the temperature at a measurement position. The identification unit compares a first threshold value determined according to the loss pattern in a first time period with a first extreme value of the absolute value of the time derivative of the temperature in the first time period, thereby identifying an anomaly corresponding to the first threshold value. [Brief explanation of the drawings]

[0006] [Figure 1] FIG. 1 is a diagram showing an example of the configuration of power electronics equipment. [Figure 2A] 10 is a graph showing an example of the time derivative value of the temperature of the copper base in the power module. [Figure 2B] FIG. 2B is a diagram showing a state when the time differential value of the temperature of the copper base in FIG. 2A is calculated. [Figure 3] An example of temperature measurements taken at 0.2 s intervals using a data logger. [Figure 4] FIG. 2 is a diagram showing an example of the functional configuration of the detection device according to the first embodiment. [Figure 5] A diagram showing an example of a Foster-type thermal network model. [Figure 6] FIG. 10 is a diagram showing an example of a temperature calculation result using a Foster-type thermal network model. [Figure 7] FIG. 1 is a diagram showing an example of a Cauer-type thermal network model. [Figure 8A] FIG. 10 shows Example 1 of the results of a temperature measurement experiment. [Figure 8B] FIG. 8B is a diagram showing the temperature measurement positions in the temperature measurement experiment of FIG. 8A. [Figure 9A] FIG. 2 shows Example 2 of the results of a temperature measurement experiment. [Figure 9B] 9B is a diagram showing measurement conditions corresponding to the respective graphs of the temperature measurement experiment of FIG. 9A. [Figure 10] FIG. 3 shows Example 3 of the results of a temperature measurement experiment. [Figure 11] 10 is a flowchart showing Example 1 of an abnormality detection process according to the first embodiment. [Figure 12] 10 is a flowchart showing a second example of the abnormality detection process according to the first embodiment. [Figure 13] FIG. 10 is a diagram showing an example of the configuration of an anomaly detection system according to a second embodiment. [Figure 14] FIG. 10 is a diagram showing an example of the configuration of an anomaly detection system according to a third embodiment. [Figure 15] FIG. 2 is a diagram showing an example of the hardware configuration of a detection device according to the first embodiment and a cloud system according to the second and third embodiments. DETAILED DESCRIPTION OF THE INVENTION

[0007] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of a detection device, an anomaly detection system, and an anomaly detection method will be described in detail with reference to the accompanying drawings.

[0008] By switching from time-based maintenance (TBM: Time Based Maintenance) to condition-based maintenance (predictive maintenance, CBM: Condition Based Maintenance), it is possible to reduce downtime, reduce labor costs, and improve customer reliability.

[0009] The causes of temperature abnormalities in power electronics equipment can be broadly divided into three categories: poor contact (for example, increased contact thermal resistance due to grease pumping out or uneven contact caused by impact), reduced cooling air flow rate (for example, clogged intake and exhaust ports or obstructions to the flow path due to obstacles), and high ambient air temperature. However, it is difficult to directly measure contact and clogging conditions using inexpensive sensors.

[0010] First, an example of a power electronics device that is a target for anomaly detection will be described.

[0011] 1 is a diagram showing an example of the configuration of a power electronics device 1. The example in Fig. 1 shows an example of the configuration of a general forced-air-cooling power electronics device 1. The power electronics device 1 includes a power module 10, a TIM 11, and a heat sink 12.

[0012] The power module 10 includes a plurality of heat-generating chips 13 and a copper base 14. The heat-generating chips 13 generate heat at a predetermined timing for each phase. For example, in the case of a three-phase AC, the heat-generating chips 13 are divided into three groups that generate heat with a phase shift of 120°. The copper base 14 is the substrate of the power module 10.

[0013] The TIM (Thermal Interface Material) 11 is a thermally conductive material for improving the thermal conductivity between components (in the example of FIG. 1, the heat sink 12 and the copper base 14). Specifically, the TIM 11 is grease, a heat dissipation sheet, or the like.

[0014] The heat sink 12 is a heat radiator. There are various methods for heat dissipation by the heat sink 12, and for example, a phase change may be used.

[0015] For example, there is a method of measuring the amount of heat generated and the temperature at multiple locations, and detecting changes in thermal resistance using the thermal circuit model shown in Figure 7, which will be described later, and applying the following formula (1). However, anomaly detection methods based on multiple sensor signals tend to be expensive and complicated. 12 is the thermal resistance of the object, Q is the amount of heat generated, and C1 is the heat capacity. T1 is the temperature on the heat source side, and T2 is the thermal resistance R 12 is the temperature via

[0016]

number

[0017] Another method is to incorporate a temperature sensor into the power module, but this method is expensive unless the power module is manufactured in-house.

[0018] Fig. 2A is a graph showing an example of the time differential value of the temperature of the copper base 14 in the power module 10. Fig. 2A shows the time differential value of the temperature of the copper base 14 in the power module 10 when an experiment was conducted in the power electronics device 1 shown in Fig. 1 by changing the state of the TIM 11 and the strength of the cooling air (presence or absence of clogging) as shown in Fig. 2B.

[0019] 2B is a diagram showing the state when the time differential value of the temperature of the copper base 14 in FIG. 2A is calculated. HS Clogging (%) indicates the percentage of clogging in the heat sink 12. That is, 0% indicates a state where the heat sink 12 is not clogged.

[0020] The time derivative value changes depending on Δt (measurement interval). Furthermore, unless Δt (measurement interval) is extremely large, the time derivative value will have multiple values, and each time derivative value will differ significantly depending on the measurement time s.

[0021] Furthermore, high-precision measurements are difficult due to the influence of noise within the power module 10. The time constant τ is also small, so the measurement interval must be short, which leads to problems such as an increase in the amount of data and the need for expensive measuring equipment.

[0022] Figure 3 shows an example of temperature measurements taken at 0.2 second intervals using a data logger. Measurements less than 1 second are generally difficult, and expensive equipment is required to remove noise.

[0023] In the following embodiment, a detection device will be described that is intended for a power electronics device 1 that is forcedly cooled on one side or naturally cooled on the other side.

[0024] (First embodiment) First, an example of the functional configuration of the detection device of the first embodiment will be described.

[0025] [Example of functional configuration] 4 is a diagram showing an example of the functional configuration of the detection device 2 of the first embodiment. The detection device 2 of the first embodiment includes a control unit 21, a storage unit 22, and an output unit .

[0026] The control unit 21 includes an acquisition unit 211 , a loss calculation unit 212 , an operation confirmation unit 213 , a time differential value calculation unit 214 , an identification unit 215 , and an output control unit 216 .

[0027] The acquisition unit 211 acquires a signal indicating the temperature from the sensor 3 placed at the measurement position (for example, below the power module 10).

[0028] The loss calculation unit 212 calculates the loss (heat generation) pattern of the power module 10 over time.

[0029] The operation confirmation unit 213 performs an operation confirmation of the power electronics device 1 that is the target of abnormality detection. For example, the power electronics device 1 that is the target of abnormality detection is a power module 10 used in an elevator, and the operation confirmation is a confirmation of the start (ON) / standby (OFF) of the power module 10.

[0030] The time differential value calculation unit 214 calculates the time differential value of the temperature at the measurement position.

[0031] The identifying unit 215 identifies an abnormality corresponding to the threshold value by comparing the threshold value determined according to the loss pattern in the specific time period with the extreme value of the absolute value of the time derivative of the temperature in the specific time period.

[0032] The abnormality corresponding to the threshold value is, for example, an abnormality in the ambient temperature of the power module 10, an abnormality at a position upstream of the measurement position in the heat conduction, and an abnormality at a position downstream of the measurement position in the heat conduction. Specifically, an abnormality at a position upstream of the measurement position in the heat conduction includes an abnormality due to poor contact of components of the power module 10. An abnormality at a position downstream of the measurement position in the heat conduction includes an abnormality (e.g., clogging) of a heat sink in contact with the power module.

[0033] The output control unit 216 notifies the user of the identified cause of the abnormality by outputting the cause of the abnormality to the output unit 23 such as a display device.

[0034] The storage unit 22 stores, for example, loss pattern information, threshold information, and time period information. The loss pattern information includes a plurality of loss patterns identified by identification information such as a loss pattern number. For example, in the case of power electronics equipment 1 installed in an elevator, the time series value of the heat generation amount is determined by the load on the car and the operation pattern, such as the number of floors traveled (travel height (m)). Therefore, the operation pattern may be used as the heat generation (loss) pattern. In other words, if the operation pattern that affects the heat generation amount is known, the heat generation amount can be known and a threshold can be set.

[0035] Many devices and systems have specific operating patterns. For example, an elevator uses a lot of power when accelerating immediately after starting operation, and then saves power during constant speed operation thereafter. Then, when the elevator decelerates, a large amount of power is applied again. Here, the period of time during which low loss continues depends on the number of floors traveled. Furthermore, the magnitude of the loss (heat generation) for high and low losses depends on the load weight (number of people). The loss pattern is the time history of the loss amount, which is determined by the number of floors traveled and the load weight, etc.

[0036] Elevators are usually designed so that the "car with passengers" and the "weight" are balanced at half the capacity weight. Therefore, the maximum power (= maximum heat generation) is generated when there are no passengers and when the elevator is full (at capacity weight). When operating with no passengers, the same current basically flows (high reproducibility of heat generation patterns). Therefore, for example, performing periodic diagnosis when operating with no passengers from a base floor on a high-rise floor to the first floor is effective for highly accurate diagnosis.

[0037] The threshold information and the time period information are determined for each loss pattern. The threshold information includes a threshold value and an abnormality corresponding to the threshold. The time period information indicates a time period in which an abnormality is determined using the threshold.

[0038] The output unit 23 is a display device or the like that outputs the cause of the abnormality identified by the identification unit 215. Note that the output unit 23 may also be a communication device that transmits the cause of the abnormality to a user's management terminal, a management server device, or the like.

[0039] The principle of anomaly detection by the detection device 2 of FIG. 4 will now be described in detail.

[0040] [Anomaly detection principle] FIG. 5 is a diagram showing an example of a Foster-type thermal network model. In the Foster-type thermal network model shown in FIG. 5, the transient thermal impedance Z at time t is calculated using the following equation (2): th (t) is calculated. Q const is the amount of heat added to node n, and is a value that does not change from time 0 to t. T nis the temperature at node n.

[0041]

number

[0042] where Z th If (t) is complex, Z th (t) can be divided into N parts along the time series and expressed as a superposition of each element. In this case, the i-th element is Z thi Let (t) be Z. thi Time constant τ at (t) i is basically Z th(i+1) Time constant τ at (t) (i+1) Greater than Z th (t) is not suitable for storage batteries and other devices with large time constants τ of components close to the heat source, but is effective for power electronics devices 1. i When a Foster-type thermal network is constructed, the transient thermal impedance of the ith element is expressed as Z thi In this case, R thi is the thermal resistance of the i-th element divided along the time series.

[0043] FIG. 6 shows an example of the temperature calculation results using the Foster-type thermal network model. thi The time period when the slope of (t) reaches its maximum is different for each node. The temperature difference between nodes becomes almost constant after all the heat has been transferred to the node in question. In power electronics equipment 1, where the time constants of each node (RC = τ, R is thermal resistance, C is heat capacity) are significantly different, the upstream node can easily reach a state of 3τ, which is 95% of the steady state, or 5τ, which is 99%, before heat is transferred to the adjacent node.

[0044] FIG. 7 is a diagram showing an example of a Cauer-type thermal network model. The mathematical model of the Cauer-type thermal network model shown in FIG. 7 is expressed by the following formula (3). T1 to T4 are the temperatures of nodes 1 to 4. R 12 is the thermal resistance between nodes 1 and 2. R 23 is the thermal resistance between nodes 2 and 3. R 34is the thermal resistance between nodes 3 and 4. C1 to C3 are the heat capacities of nodes 1 to 3.

[0045]

number

[0046] Thermal resistance R 23 Here, we will explain the case where a change in temperature (such as grease deterioration) is detected by the downstream temperature T3 on the heat dissipation path. Here, the following formula (4) holds true in the short time period from when heat is transmitted to node 2 and the temperature of node 3 begins to rise.

[0047]

number

[0048] Substituting the above equation (4) into the following equation (5), which is obtained by further differentiating dT3 / dt in the above equation (3), and determining the condition under which the value obtained by differentiating T3 twice is 0, yields the following equation (6).

[0049]

number

[0050]

number

[0051] R 23 As this changes, the time when the second derivative of T3 becomes 0 also changes. However, in many cases (the time when the second derivative of T3 becomes 0) >> (the time constant of node T2), and T1-T2 is almost constant (95% for 3τ, 99% for 5τ).

[0052] Therefore, even if the time at which the value obtained by differentiating T3 twice becomes 0 changes, T1-T2 itself is hardly affected.

[0053] In other words, R 23When becomes large, from the above equation (6), dT3 / dt becomes R 34 / (R 23 +R 34 ) is smaller. From a practical standpoint, it is important to place the temperature measurement point near the thermal resistance part of the object whose condition is to be detected.

[0054] By modifying the above formula (3), the following formula (7) is obtained.

[0055]

number

[0056] If it is assumed that the values ​​obtained by differentiating T1, T2, and T3 once are approximately equal, then the above formula (7) can be transformed into the following formula (8).

[0057]

number

[0058] By generalizing equation (8), the following equation (9) is obtained.

[0059]

number

[0060] From the above equation (9), it can be seen that the temperature differential value when heat is transmitted to a certain node n depends on the following equation (10).

[0061]

number

[0062] In the power electronics device 1, the heat capacity C and the heat resistance R often increase from upstream to downstream.

[0063] So, for example, R 23 When detecting the change in dT6 / dt, the following equation (11) is obtained:23 The change in R 23 The denominator is so large that the change in

[0064]

number

[0065] Although an actual system (power electronics device 1) is not as simple as the example of the Cauer thermal network model in FIG. 7, it shows the same tendency as explained in the above description of FIG. 7. In the example of FIG. 1, the most upstream is the heat-generating chip 13, and the most downstream is the heat sink 12. The heat-generating chip 13 in FIG. 1 corresponds to node T1 in the Cauer thermal network model in FIG. 7. n The larger n is, the more downstream the node is.

[0066] The above explanation was about the heating state (change from heat generation ON), but the same tendency is observed in the cooling state. In the cooling state, the sign of dT3 / dt is negative.

[0067] Therefore, R 23 As the value of dT3 / dt increases, the absolute value of the peak value of dT3 / dt decreases.

[0068] Downstream anomalies near the temperature measurement point, e.g., R 34 The same can be said about the increase in R 34 As this increases, the peak value of dT3 / dt increases.

[0069] Abnormalities at a position thermally distant from the temperature measurement position can be determined by considering the system as a lumped capacitance system and differentiating the above equation (2) once, as given by the following equation (12):

[0070]

number

[0071] Equation (12) above shows that an increase in the time constant τ increases the time it takes to reach a steady state. Therefore, by comparing the magnitude of the absolute value of dT / dt during the time period corresponding to the time constant τ when considered as a lumped capacitance system, it is possible to determine an abnormality at the most downstream side (such as a decrease in wind speed).

[0072] With the time constant τ in the above equation (12) as the reference, 3τ and 5τ are the times to reach 95% and 99%, respectively, of the steady state.

[0073] When checking the operation of the power electronics device 1, if a predetermined time, for example, 5τ or more, has passed since the end of the previous operation, the temperature of the sensor 3 at the start of operation is approximately equal to the air temperature. Therefore, an abnormality in the air temperature can be determined.

[0074] In addition, when the cooling equipment is stopped when the operation is finished, a time constant τ based on the cooling capacity during the stoppage is used.

[0075] [Verification of effectiveness] Fig. 8A is a diagram showing Example 1 of the results of a temperature measurement experiment. Fig. 8A shows the results of a temperature measurement experiment of an inverter equipped with a power module 10. The state (experimental conditions) when calculating the time differential value of the temperature of the copper base 14 in Fig. 8A is the same as that in Fig. 2B.

[0076] Fig. 8B is a diagram showing the temperature measurement positions in the temperature measurement experiment of Fig. 8A. The power module 10 is attached to the heat sink 12, and the sensor 3 is placed in an area where the shadow of the power module 10 is cast when the power module 10 is projected vertically toward the heat sink 12, and the sensor 3 acquires the temperature at the measurement position in the figure.

[0077] 8B, the temperature of the heat sink 12 directly below the power module 10 is measured. The state of the thermal resistance between the power module 10 and the heat sink 12 depends on the state of the TIM 11 (for example, thermal grease).

[0078] When detecting a change in the thermal resistance between the power module 10 and the heat sink 12, it is desirable to use the temperature of the heat sink 12 directly below the power module 10, which is in the vicinity of the power module 10 (see FIG. 8B).

[0079] It should be noted that measuring the temperature inside the power module 10, which is located upstream on the heat dissipation path, tends to be expensive.

[0080] The time constant at the measurement position in Figure 8B is on the order of 1 second, and there is no significant difference in the extreme absolute value of the time derivative of the temperature whether the measurement interval is 1 second or 2 seconds. Note that in a typical power module 10, heat is transferred to the case in the order of 0.1 seconds. This can be seen by looking at the transient thermal resistance curve available from specifications, etc.

[0081] Being able to extend the temperature measurement interval has advantages in terms of noise and data volume. For example, elevators often operate for several tens of seconds, so a value of 1 second is not too large. The detection device 2 of the first embodiment is provided with a determination time period. This allows abnormalities to be detected without being affected by the conditions of locations with different orders of time constant τ.

[0082] The experimental conditions in Fig. 8A (see Fig. 2B) include the heat sink clogging state, air temperature, etc. The experimental conditions in Fig. 8A (see Fig. 2B) also include an experiment in which the initial component temperature is not equal to the air temperature.

[0083] Naturally, heat sink clogging and air temperature affect the steady-state temperature. Using this embodiment, which sets a time period for judgment, it can be seen from FIG. 8A that the effects of heat sink clogging and air temperature can be eliminated. By changing the time period for judgment, it becomes possible to detect anomalies in multiple locations and identify their causes.

[0084] Fig. 9A is a diagram showing Example 2 of the results of the temperature measurement experiment, and Fig. 9B is a diagram showing measurement states corresponding to the graphs of the temperature measurement experiment in Fig. 9A.

[0085] Figure 9A shows the results of the same experiment as Figure 8A, but the time period of interest is changed. The clogged state of the heat sink 12 is located at a position thermally distant from the heat sink 12 directly below the power module 10, which is the temperature measurement point.

[0086] Figure 9A shows that the time derivative of temperature increased due to the increase in thermal resistance. This can also be seen from equation (7) above. Figure 9A also shows that other conditions, such as the state of the grease, had no effect.

[0087] In the example of Fig. 9A, when time period B (second time period) is set to 200 to 600 s, the minimum value of the time derivative of the temperature is 0.02 if the clogging of the heat sink 12 is 0%. Also, when the clogging of the heat sink 12 is 100%, the minimum value is 0.035. When the threshold value for determining the minimum value of the time derivative of the temperature is set to 0.03, a state in which the clogging of the heat sink 12 is about 75% is identified.

[0088] 2A and 2B show the time derivatives of temperature when the temperature of the copper base 14 in the power module 10 is used. When the time derivatives of temperature are compared at a measurement interval of 1 s, the maximum absolute value differs for each experimental condition, making it difficult to determine the state. Because the time constant τ of the position of the copper base 14 in the power module 10 is small, it is difficult to determine the extreme absolute value of the time derivative of temperature unless the measurement interval is set to less than 1 s.

[0089] Fig. 10 is a diagram showing Example 3 of the results of a temperature measurement experiment. The example in Fig. 10 shows the results when using the temperature of the heat sink 12 near the power module 10. The state (experimental conditions) when calculating the time derivative of the temperature in Fig. 10 is the same as Fig. 2B.

[0090] In the case of the temperature of a heat sink that is not directly below the power module 10, the maximum absolute value of the time derivative of the temperature is small, which indicates that the time constant is large at a position that is not directly below the power module 10. This corresponds to the situation of the following equation (13) in the above equation (9).

[0091]

number

[0092] From this, R 23 At a thermally distant location, R 23 It is clear that it is difficult to detect this state.

[0093] Based on the principle of anomaly detection described above, the detection device 2 of the first embodiment detects the state of the detection target using the time differential value of the signal from the sensor 3, which measures the temperature and is located downstream on the heat dissipation path from the position of the detection target whose state is being detected. Specifically, the detection device 2 detects the state of the detection target by comparing the extreme value of the absolute value of the time differential value in a predetermined determination time period with a threshold. In this specification, the extreme value is explained as a maximum value or a minimum value within a specified range.

[0094] The predetermined determination time period is set based on the elapsed time from the heating or cooling start time, taking into consideration the thermal time constant τ at the temperature measurement position, the position of the state detection target, etc. Specifically, the heating or cooling start time is the start or end time of operation of the power module 10.

[0095] As shown in equation (6) above, the extreme value of the absolute value of the time derivative depends on the amount of heat generated (loss), so the threshold value is changed depending on the heat generation (loss) pattern (operating pattern). As shown in equation (6), the time derivative of temperature depends on the magnitude of Q and the magnitude of thermal resistance. If Q is kept constant, the time derivative of temperature depends only on thermal resistance. As shown in equation (3) above, this thermal resistance also affects the temperature of upstream components. Therefore, the relationship between thermal resistance and temperature according to the amount of heat generated is organized, and the time derivative value under which the temperature rise due to an increase in thermal resistance exceeds a specified value is set as the threshold. For example, the temperature time derivative value at which the chip temperature in a power module rises 10°C above normal during a certain heat generation pattern is set as the threshold. This can also be determined experimentally.

[0096] It is not necessary to know the time series values ​​of the heat generation amount in detail. For example, the time series values ​​of the heat generation amount can be replaced by the output current or the like.

[0097] When the temperature of the heat sink 12 directly below the power module 10 is used, the sensor 3 may be installed by drilling, machining, or cutting a hole in the base of the heat sink 12 .

[0098] There may be multiple sensors 3. For example, if there are multiple sensors 3 in a small area, the condition may be determined comprehensively from the analysis results of the signals from each sensor 3. For example, if there are multiple sensors 3 in a wide area, the location (factor) for grasping the condition may be divided for each sensor 3.

[0099] Furthermore, instead of a time differential value, a difference value or an integral value between two points may be used.

[0100] [Example of anomaly detection method] 11 is a flowchart showing Example 1 of the abnormality detection process of the first embodiment. First, the identifying unit 215 sets a specific phenomenon (step S1). For example, the specific phenomenon is a contact failure of a component of the power module 10.

[0101] Next, the operation check unit 213 checks the operation of the power electronics device 1 that is the target of abnormality detection (step S2).

[0102] Next, the acquisition unit 211 acquires a sensor signal from the sensor 3 (step S3).

[0103] Next, the loss calculation unit 212 calculates, for example, a loss (heat generation) pattern of the power module 10 over time (step S4).

[0104] Next, the identification unit 215 determines whether the loss pattern calculated in step S4 is a loss pattern designated in the detection device 2 (step S5). The designated loss pattern is a loss pattern stored as loss pattern information in the storage unit 22. If the loss pattern is not a designated loss pattern (step S5, No), the anomaly detection process ends.

[0105] If the loss pattern is the specified one (step S5, Yes), the identifying unit 215 assigns a loss pattern number to the loss pattern calculated in step S4, and acquires the first and second thresholds associated with the loss pattern number and the time period A (first time period) from the storage unit 22 (step S6).

[0106] Next, the time differential value calculation unit 214 calculates the time differential value dT / dt of the temperature T (step S7).

[0107] In the following steps S8 and S9, an abnormality in the change in temperature T over time is determined based on the absolute value of the time derivative dT / dt.

[0108] First, the identification unit 215 determines whether the maximum absolute value of the time derivative dT / dt in time period A is greater than a first threshold value (step S8). The maximum absolute value of the time derivative dT / dt is the time derivative dT / dt at time t at which the value obtained by differentiating dT / dt again becomes 0.

[0109] Next, if the maximum absolute value of the time derivative dT / dt in time zone A is greater than the first threshold (step S8, Yes), the identification unit 215 determines whether the time showing the maximum absolute value of the time derivative dT / dt in time zone A is smaller than the second threshold (step S9).

[0110] If the time showing the maximum absolute value of the time derivative value dT / dt in time zone A is smaller than the second threshold value (step S9, Yes), the identification unit 215 determines that no abnormality has occurred at a position upstream of the measurement position in the heat conduction, and terminates the abnormality detection process.

[0111] On the other hand, if the maximum absolute value of the time derivative dT / dt in time zone A is less than or equal to the first threshold (step S8, No), and if the time showing the maximum absolute value of the time derivative dT / dt in time zone A is greater than or equal to the second threshold (step S9, No), the identification unit 215 identifies the cause of the abnormality (for example, poor contact of components of the power module 10) (step S10).

[0112] Next, the output control unit 216 notifies the user of the cause of the abnormality by outputting the cause of the abnormality identified in step S10 to the output unit 23 such as a display device (step S11).

[0113] Fig. 12 is a flowchart showing Example 2 of the abnormality detection process of the first embodiment. In the example of Fig. 12, determinations based on a third threshold and a fourth threshold are added. The determination based on the third threshold is an abnormality determination for the ambient temperature of the power module 10. The determination based on the fourth threshold is an abnormality determination at a position downstream of the measurement position in the heat conduction direction.

[0114] The explanation of steps S21 to S23 is omitted because it is the same as the explanation of steps S1 to S3 in FIG.

[0115] The identifying unit 215 determines whether a predetermined time has elapsed since the previous operation of the power module 10 (step S24). If the predetermined time has elapsed (step S24, Yes), the identifying unit 215 determines whether the temperature indicated by the sensor signal acquired in step S23 is lower than a fourth threshold value (air temperature threshold value) (step S25).

[0116] If the temperature indicated by the sensor signal is equal to or higher than the fourth threshold value (step S25, No), the identifying unit 215 identifies an abnormality in the ambient temperature of the power module 10 as the cause of the abnormality (step S33). The process of step S34 is the same as step S11 in FIG. 11, and therefore a description thereof will be omitted.

[0117] If the temperature indicated by the sensor signal is lower than the fourth threshold (Yes in step S25), the abnormality detection process proceeds to step S26. The processes in steps S26 and S27 are similar to steps S4 and S5 in FIG. 11, and therefore will not be described.

[0118] Next, the identifying unit 215 assigns a loss pattern number to the loss pattern calculated in step S26, and acquires the first to third thresholds and time periods A and B associated with the loss pattern number from the storage unit 22 (step S28).

[0119] The processing in steps S29 to S31 is the same as steps S7 to S9 in FIG. 11, and therefore a description thereof will be omitted.

[0120] Next, if the time showing the maximum absolute value of the time derivative dT / dt in time zone A is smaller than the second threshold (step S31, Yes), the identification unit 215 determines whether the minimum absolute value of the time derivative dT / dt in time zone B is smaller than the third threshold (step S32).

[0121] If the maximum absolute value of the time differential value dT / dt in time zone B is equal to or greater than the third threshold value (step S32, No), the identification unit 215 identifies the cause of the abnormality as an abnormality that occurred at a position downstream of the measurement position in the heat conduction direction (for example, an abnormality in the heat sink in contact with the power module) (step S33).

[0122] On the other hand, if the absolute value of the time differential value dT / dt in time zone B is smaller than the third threshold value (step S32, Yes), the identification unit 215 determines that no abnormality has occurred in the ambient temperature of the power module 10, at a position upstream of the measurement position in terms of heat conduction, or at a position downstream of the measurement position in terms of heat conduction, and terminates the abnormality detection process.

[0123] As described above, the detection device 2 of the first embodiment can detect the state of the operating environment of the power module 10 with a simpler configuration. For example, an abnormality in the operating environment of the power module 10 can be detected with a configuration including the detection device 2 shown in Fig. 4 and the sensor 3 arranged at the measurement position shown in Fig. 8B.

[0124] (Second embodiment) Next, a second embodiment will be described. In the description of the second embodiment, the same description as in the first embodiment will be omitted, and only the differences from the first embodiment will be described.

[0125] [Configuration example] 13 is a diagram showing an example of the configuration of an anomaly detection system 100 according to the second embodiment. The anomaly detection system 100 according to the second embodiment includes a sensor 3, a device 4, and a cloud system 5.

[0126] The description of the sensor 3 is omitted since it is the same as that of the first embodiment.

[0127] The device 4 is, for example, an elevator. The device 4 includes an acquisition unit 41, a loss calculation unit 42, an operation confirmation unit 43, and an output unit 44. The explanation of the acquisition unit 41, the loss calculation unit 42, and the operation confirmation unit 43 is omitted because it is the same as the explanation of the acquisition unit 211, the loss calculation unit 212, and the operation confirmation unit 213 in the first embodiment.

[0128] The output unit 23 outputs the signal of the sensor 3 acquired by the device 4 and the loss pattern calculated by the loss calculation unit 42 to a cloud system 5 connected via a network using a wireless or wired communication device.

[0129] The cloud system 5 includes a control unit 51, a storage unit 52, and an output unit 53. The control unit 51 includes an acquisition unit 511, a time differential value calculation unit 512, an identification unit 513, and an output control unit 514.

[0130] The acquisition unit 511 acquires the signal of the sensor 3 acquired by the device 4 and the loss pattern calculated by the device 4 via a network using a wireless or wired communication device.

[0131] The explanations of the time differential value calculation unit 512, the identification unit 513, and the output control unit 514 are omitted because they are similar to the explanations of the time differential value calculation unit 214, the identification unit 215, and the output control unit 216 in the first embodiment. In addition, the explanations of the storage unit 52 and the output unit 53 are also omitted because they are similar to the explanations of the storage unit 22 and the output unit 23 in the first embodiment.

[0132] As described above, the anomaly detection system 100 of the second embodiment includes the sensor 3 that acquires the temperature at a measurement position, the device 4, and the cloud system 5. In the device 4, the acquisition unit 41 acquires the temperature from the sensor 3. The loss calculation unit 42 calculates a loss pattern of the power module 10 over time. Then, the output unit 44 outputs the temperature and the loss pattern to the cloud system 5. In the cloud system 5, the time derivative calculation unit 512 calculates a time derivative of the temperature. Then, the identification unit 513 identifies an anomaly corresponding to the first threshold by, for example, comparing a first threshold determined according to the loss pattern for time period A with the extreme value of the absolute value of the time derivative of the temperature for time period A.

[0133] That is, in the second embodiment, the signal from the sensor 3 is processed on the device 4 (edge ​​device) side, and the signal from the sensor 3 and the loss pattern obtained by the device 4 are collected in the cloud system 5. Note that although the example in Fig. 13 shows a case where there is one device 4, there may be multiple devices 4 communicating with the cloud system 5.

[0134] According to the second embodiment, data used for detecting anomalies is acquired on the side of one or more edge devices, and anomaly detection can be performed in an integrated manner on the side of the cloud system 5.

[0135] The data to be transmitted to the cloud system 5 may be obtained by other methods. For example, the data used to determine an abnormality may be obtained via email, a USB memory, or the like.

[0136] (Third embodiment) Next, a third embodiment will be described. In the description of the third embodiment, the same description as in the first and second embodiments will be omitted, and only differences from the first and second embodiments will be described. In the third embodiment, a case will be described in which more data is processed on the edge device side than in the second embodiment.

[0137] [Configuration example] 14 is a diagram showing an example of the configuration of an anomaly detection system 100-2 according to the third embodiment. The anomaly detection system 100-2 according to the third embodiment includes a detection device 2, a sensor 3, and a cloud system 5-2.

[0138] The detection device 2 and the sensor 3 will be described below as they are the same as those in the first embodiment. In the third embodiment, the output unit 23 outputs the detection result by the detection device 2 to the cloud system 5-2 connected via a network using a wireless or wired communication device.

[0139] The detection results include, for example, the signal of the sensor 3 acquired by the acquisition unit 211, the loss pattern calculated by the loss calculation unit 212, the time differential value calculated by the time differential value calculation unit 214, and the abnormality identified by the identification unit 215.

[0140] The cloud system 5-2 includes an acquisition unit 54 and a control unit 55.

[0141] The acquisition unit 54 acquires the above-described detection results obtained by the detection device 2 via a network using a wireless or wired communication device. The acquisition unit 54 acquires the detection results from the detection device 2 in real time.

[0142] Although the example in FIG. 14 shows a case where there is one detection device 2, there may be a plurality of detection devices 2 communicating with the cloud system 5-2.

[0143] The control unit 55 performs various processes based on the detection results. For example, the control unit 55 determines the timing of maintenance of the elevator equipped with the power electronics device 1 based on the detection results acquired in real time.

[0144] By using detection results obtained in real time, it is possible to reduce, for example, unnecessary maintenance. Here, unnecessary maintenance refers to time-based maintenance (TBM). Specifically, reducing unnecessary maintenance can prevent the replacement of parts that are still usable. Replacing parts that are still usable wastes time and environmental costs, and leads to the risk of early defects, so reducing unnecessary maintenance is beneficial.

[0145] For example, by using detection results obtained in real time, initial action in the event of an abnormality can be taken more quickly than if the customer operating the elevator were to become aware of the abnormality and then contact the elevator.

[0146] Furthermore, the control unit 62 can make a qualitative judgment by comparing the detection results of various power electronics devices 1 obtained from multiple detection devices 2. Furthermore, the control unit 62 continues to record the state when a problem occurs, thereby obtaining a correlation between the probabilistic failures and the state of the power electronics devices 1.

[0147] The state of the power electronics device 1 may include the deterioration state and remaining lifespan predicted based on the temperature, in addition to the cooling performance of the power electronics device 1. For example, the control unit 62 can qualitatively compare how long it will take for the frequency of abnormality occurrence to increase by comparing the detection results of multiple power electronics devices 1 that are operating with the same cooling performance (similar usage).

[0148] Finally, examples of the hardware configuration of the detection device 2 of the first embodiment and the cloud systems 5 and 5-2 of the second and third embodiments will be described.

[0149] [Example of hardware configuration] 15 is a diagram illustrating an example of the hardware configuration of the detection device 2 of the first embodiment and the cloud systems 5 and 5-2 of the second and third embodiments. The detection device 2 and the cloud systems 5 and 5-2 include a processor 201, a main storage device 202, an auxiliary storage device 203, a display device 204, an input device 205, and a communication device 206. The processor 201, the main storage device 202, the auxiliary storage device 203, the display device 204, the input device 205, and the communication device 206 are connected via a bus 210.

[0150] The detection device 2 and the cloud systems 5 and 5-2 may not be provided with some of the above configurations. For example, if the detection device 2 and the cloud systems 5 and 5-2 can use the input function and display function of an external device, the detection device 2 and the cloud systems 5 and 5-2 may not be provided with the display device 204 and the input device 205.

[0151] The processor 201 executes a program read from the auxiliary storage device 203 to the main storage device 202. The main storage device 202 is a memory such as a ROM and a RAM. The auxiliary storage device 203 is a HDD, a memory card, or the like.

[0152] The display device 204 is, for example, a liquid crystal display. The input device 205 is an interface for operating the detection device 2 and the cloud systems 5 and 5-2. The display device 204 and the input device 205 may be realized by a touch panel or the like having a display function and an input function. The communication device 206 is an interface for communicating with other devices.

[0153] For example, the detection device 2 and the programs executed by the cloud systems 5 and 5-2 are provided as computer program products in the form of installable or executable files recorded on computer-readable storage media such as memory cards, hard disks, CD-RWs, CD-ROMs, CD-Rs, DVD-RAMs and DVD-Rs.

[0154] Furthermore, for example, the detection device 2 and the programs executed by the cloud systems 5 and 5-2 may be stored on a computer connected to a network such as the Internet and provided by being downloaded via the network.

[0155] Furthermore, for example, the programs for the detection device 2 and the cloud systems 5 and 5-2 may be configured to be provided in a state where they are pre-installed in a ROM or the like.

[0156] The detection device 2 and the programs executed by the cloud systems 5 and 5-2 have a modular configuration that includes functions that can be realized by programs among the above-mentioned functional configurations. As for each function, the processor 201 reads and executes the program from a storage medium, and the above-mentioned functional blocks are loaded onto the main memory device 202 as actual hardware. In other words, the above-mentioned functional blocks are generated on the main memory device 202.

[0157] Note that some or all of the above-described functions may be realized by hardware such as an integrated circuit (IC) rather than by software.

[0158] Furthermore, each function may be realized using a plurality of processors 201, in which case each processor 201 may realize one of the functions, or may realize two or more of the functions.

[0159] Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, and are also included in the scope of the invention and its equivalents as defined in the claims. [Explanation of symbols]

[0160] 1 Power electronics equipment 2. Detection device 3 sensors 4 equipment 5. Cloud System 10 Power Module 11 TIM 12 Heat sink 13 Heating Chip 14 Copper base 21 Control Unit 22 Memory section 23 Output section 41 Acquisition Department 42 Loss calculation section 43 Operation Check Section 44 Output section 51 Control section 52 Storage section 53 Output section 54 Acquisition Department 55 Control Unit 100 Anomaly Detection System 201 processor 202 Main storage 203 Auxiliary storage device 204 Display device 205 Input Device 206 Communication Equipment 210 Bus 211 Acquisition Department 212 Loss calculation section 213 Operation Check Department 214 Time derivative calculation unit 215 Specific section 216 Output control section 511 Acquisition Department 512 Time derivative calculation unit 513 Specific part 514 Output control section

Claims

1. a loss calculation unit that calculates a loss pattern of the power module over time; a time differential value calculation unit that calculates a time differential value of the temperature at the measurement position; an identification unit that identifies a state corresponding to a first threshold value by comparing a first threshold value determined in accordance with the loss pattern in a first time period with a first extreme value of an absolute value of a time derivative of the temperature in the first time period; A detection device comprising:

2. the power module is attached to a heat sink; a sensor is disposed in an area where a shadow of the power module is cast when the power module is vertically projected toward the heat sink; The sensor acquires the temperature at the measurement location. The detection device of claim 1 .

3. When a predetermined time has elapsed since the end of the previous operation of the power module, the identification unit acquires the temperature acquired from the sensor as the ambient air temperature of the power module, and further identifies the ambient air temperature by comparing the ambient air temperature with an air temperature threshold. The detection device of claim 2 .

4. the first extremum is the maximum value of the time derivative, the identifying unit identifies a state of a position upstream of the measurement position in the thermal conduction direction when the first extreme value is equal to or less than the first threshold value. The detection device according to any one of claims 1 to 3.

5. The condition at the upstream position includes an abnormality due to poor contact of a component of the power module. The detection device according to claim 4.

6. the identifying unit further compares the time at which the time differential value of the temperature reaches the first extreme value with a second threshold value, thereby further identifying the state corresponding to the second threshold value. The detection device according to any one of claims 1 to 3.

7. the identifying unit further identifies a state corresponding to the third threshold by comparing a third threshold determined according to the loss pattern in a second time period that is later than the first time period with a second extreme value of the absolute value of the time derivative of the temperature. The detection device according to any one of claims 1 to 3.

8. the second extremum is a minimum value of the time differential value, the identifying unit identifies a state of a position downstream of the measurement position in the heat conduction direction when the second extreme value is equal to or greater than the third threshold value. The detection device of claim 7.

9. the downstream condition includes an abnormality in a heat sink in contact with the power module; The detection device of claim 8.

10. The first and second time periods are determined based on a thermal time constant at the measurement location. The detection device of claim 7.

11. The first and second time periods are determined based on a start time or an end time of operation of the power module. The detection device of claim 7.

12. a sensor for acquiring a temperature at a measurement location; The device, A cloud system, The device comprises: an acquisition unit that acquires the temperature from the sensor; a loss calculation unit that calculates a loss pattern of the power module over time; an output unit that outputs the temperature and the loss pattern to the cloud system, The cloud system includes: a time differential value calculation unit that calculates a time differential value of the temperature; an identification unit that identifies a state corresponding to a first threshold value by comparing a first threshold value determined in accordance with the loss pattern in a first time period with a first extreme value of an absolute value of a time derivative of the temperature in the first time period; A detection system comprising:

13. a detection device calculating a loss pattern of the power module over time; a step in which the detection device calculates a time derivative value of the temperature at the measurement position; The detection device compares a first threshold determined according to the loss pattern in a first time period with a first extreme value of the absolute value of the time derivative of the temperature in the first time period, thereby identifying a state corresponding to the first threshold; A detection method comprising:

Images