Detection device, detection system, and computer program

Abstract

Providing detection devices, detection systems, and computer programs. [Solution] The system includes an acquisition unit that acquires measurement data obtained by measuring the induced current and induced voltage induced by a reactor connected to the power conversion circuit portion of a power supply device in a time series, a calculation unit that calculates the amount of change in impedance in the reactor over time based on the acquired measurement data, and a detection unit that detects anomalies in the reactor based on the calculated amount of change in impedance.

Description

Technical Field

[0001] The present disclosure relates to a detection device, a detection system, and a computer program.

Background Art

[0002] A reactor (inductor or transformer) is connected to the power conversion circuit of a power supply device. By appropriately controlling the current flowing through the reactor, the power supply device can function as a coupling inverter that couples the power conversion circuit to the power grid (see, for example, Patent Documents 1 and 2).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] The reactor used in a power supply device may experience aging deterioration due to its own heat generation, ambient temperature, stress applied to the reactor, etc., and may suddenly occur as an abnormality.

[0005] However, for a power supply device that is actually operating, there is no effective diagnostic technique that can detect an abnormality in real time according to the operation of the power conversion circuit. Therefore, after the power supply device fails, the failure of the reactor is grasped, and it is difficult to establish an appropriate maintenance plan.

[0006] An object of the present disclosure is to provide a detection device, a detection system, and a computer program that can detect the abnormality of a reactor connected to a power conversion circuit portion of a power supply device and be useful for a maintenance plan.

Means for Solving the Problems

[0007] A detection device according to one aspect of the present disclosure includes an acquisition unit that acquires measurement data obtained by measuring the induced current and induced voltage induced by a reactor connected to the power conversion circuit portion of a power supply device in a time series, a calculation unit that calculates the amount of change in impedance in the reactor over time based on the acquired measurement data, and a detection unit that detects anomalies in the reactor based on the calculated amount of change in impedance. [Effects of the Invention]

[0008] According to the above embodiment, it is possible to detect anomalies in the reactor connected to the power conversion circuit portion of the power supply equipment and use this information for maintenance planning. [Brief explanation of the drawing]

[0009] [Figure 1] This is a schematic diagram showing an example configuration of a remote diagnostic system according to Embodiment 1. [Figure 2] This is a block diagram showing the internal configuration of a remote diagnostic device. [Figure 3] This is a flowchart illustrating the steps performed by the remote diagnostic device. [Figure 4] This is a schematic diagram showing an example of the output of the detection results. [Figure 5] This is a schematic diagram showing an example of a learning model configuration. [Figure 6] This is a graph showing the change in probability over time. [Modes for carrying out the invention]

[0010] (1) The detection device of the present disclosure includes an acquisition unit that acquires measurement data obtained by measuring the induced current and induced voltage induced by a reactor connected to the power conversion circuit portion of a power supply device in a time series, a calculation unit that calculates the amount of change in impedance in the reactor over time based on the acquired measurement data, and a detection unit that detects anomalies in the reactor based on the calculated amount of change in impedance.

[0011] Reactors used in the power conversion circuit of power supply equipment deteriorate over time due to the heat generated by the reactor itself, ambient temperature, and stress applied to the reactor, which can suddenly manifest as malfunctions.

[0012] However, there are no effective diagnostic techniques for power supply equipment that is actually in operation. As a result, reactor failures are only diagnosed after the power supply equipment has failed. Maintenance and replacement are generally carried out based on reliability engineering bathtub curves (failure rate curves), but this does not take into account the actual operating conditions of the power supply equipment, making it difficult to create an appropriate maintenance plan. Replacing reactors used in the power conversion circuit is a very large-scale operation because it must be done on-site. In some cases, the entire power supply equipment may need to be replaced, resulting in significant economic losses. From a maintenance perspective, it is preferable to replace them in advance through periodic inspections.

[0013] According to the detection device described in (1) above, the induced current and induced voltage generated by the reactor are measured over time from power supply equipment that is actually in operation, and abnormalities in the reactor are detected based on the obtained measurement data. Therefore, the detection device can continuously monitor the state of the reactor over the period leading up to the reactor becoming abnormal, which can be useful for maintenance planning of power supply equipment.

[0014] (2) The detection device described in (1) above may be equipped with a learning unit that learns the criteria for determining when detecting an abnormality in the reactor.

[0015] According to the detection device described in (2) above, the criteria for detecting abnormalities in the reactor are determined by learning. The detection device acquires measurement data of induced current and induced voltage measured over time, so it can collect a large amount of measurement data. By determining the criteria based on a large amount of measurement data, the detection device can improve the estimation accuracy when detecting abnormalities. The detection device may also determine whether the reactor is abnormal based on the detection results from the detection unit.

[0016] (3) In the detection device according to the above (1) or (2), an estimation unit may be provided that estimates a time-series change in the probability of an abnormality occurring in the reactor using the measurement data, and estimates the life of the reactor based on the estimated time-series change.

[0017] According to the detection device described in the above (3), by estimating the time-series change in the probability of an abnormality occurring in the reactor, the life of the reactor can be estimated. As a result, the detection device can prompt the user to perform maintenance work on the power supply device including replacement of the reactor before the reactor reaches the end of its life.

[0018] (4) In the detection device according to the above (3), a reception unit that receives a user's determination as to whether to allow the abnormality detected by the detection unit, and a presentation unit that presents the user with the maintenance replacement timing of the reactor according to the user's determination result may be provided.

[0019] According to the detection device described in the above (4), the maintenance replacement timing determined according to the user's determination as to whether the abnormality of the reactor can be allowed can be presented to the user.

[0020] (5) The detection system of the present disclosure includes a power supply device including a power conversion circuit, a reactor connected to the power conversion circuit, a measurement unit that measures the induced current and induced voltage induced by the reactor in time series, an acquisition unit that acquires the measurement data of the induced current and induced voltage measured by the measurement unit from the power supply device via a communication network, a calculation unit that calculates the temporal change amount of the impedance in the reactor based on the acquired measurement data, and a detection unit that detects the abnormality of the reactor based on the calculated change amount of the impedance.

[0021] According to the detection system described in (5) above, the detection device measures the induced current and induced voltage from the reactor in a time series from the power supply equipment that is actually in operation, and detects abnormalities in the reactor based on the obtained measurement data. Therefore, the detection device can keep track of the state of the reactor in real time until the reactor becomes abnormal, which can be useful for maintenance planning of the power supply equipment.

[0022] (6) The computer program of the present disclosure acquires measurement data obtained by measuring the induced current and induced voltage induced by a reactor connected to the power conversion circuit portion of a power supply device in a time series, calculates the amount of change in impedance in the reactor based on the acquired measurement data, and causes the computer to perform a process to detect the reactor anomaly based on the calculated amount of change in impedance.

[0023] The computer program described in (6) above measures the induced current and induced voltage from the reactor in a time series from the power supply equipment that is actually in operation, and detects abnormalities in the reactor based on the obtained measurement data. Therefore, the computer can keep track of the reactor's condition in real time until it becomes abnormal, which can be useful for planning the maintenance of the power supply equipment.

[0024] The present invention will be described in detail below with reference to the drawings illustrating its embodiments. (Embodiment 1) Figure 1 is a schematic diagram showing an example configuration of a remote diagnostic system according to Embodiment 1. The remote diagnostic system 1 according to Embodiment comprises a power supply device 100 to be diagnosed and a remote diagnostic device 200 for diagnosing the power supply device 100 from a remote location. Here, the remote location represents a location far from the power supply device 100. The remote location does not necessarily have to be a location geographically distant from the power supply device 100; it is sufficient if it is a location far enough away that the power supply device 100 cannot be directly operated. The remote diagnostic system 1 may further include a user terminal 300 to which the diagnostic results from the remote diagnostic device 200 are notified.

[0025] The power supply unit 100 comprises a power conversion circuit 110, a control unit 120, and a communication unit 130. The power supply unit 100 is, for example, a grid-connected inverter connected between a DC power supply V and a power grid E. The DC power supply V is a storage battery, a solar cell, etc. The power supply unit 100 converts the DC power supplied from the DC power supply V into AC power using the power conversion circuit 110 and supplies it to the power grid E. Alternatively, the power supply unit 100 may be a rectifier, an uninterruptible power supply (UPS), an energy storage system (ESS), a power conditioner, or a V2X (Vehicle to Load, Home, Grid, etc.) including an electric vehicle power conditioner. The power supply destination of the power supply unit 100 is not limited to the power grid E, but may be any device or equipment, and may be a load such as a motor.

[0026] The power conversion circuit 110 comprises an inverter circuit 111, a reactor 112, a capacitor 113, and a measurement unit 114. The inverter circuit 111 is a circuit that converts a DC voltage input from a DC power supply V into an AC voltage having a desired frequency and voltage. The inverter circuit 111 includes, for example, a full-bridge circuit composed of multiple power semiconductor elements. The reactor 112 and capacitor 113 constitute an LC filter circuit and have the function of shaping the pulsed waveform output from the inverter circuit 111 into a sinusoidal waveform. The power conversion circuit 110 is connected to the power system E via the LC filter circuit consisting of the reactor 112 and capacitor 113 and outputs power to the system E.

[0027] The measurement unit 114 is equipped with a current sensor and a voltage sensor for measuring the induced current and induced voltage induced by the reactor 112. The measurement unit 114 measures the induced current and induced voltage induced by the reactor 112 in a time series using the current sensor and the voltage sensor, and outputs the obtained measurement data to the control unit 120. Here, the voltage sensor may measure the induced voltage non-contact using a closed coil (differential coil) as shown in Figure 1. Alternatively, the induced current may be calculated from the measured induced voltage using Ohm's law by utilizing a shunt resistor or the like provided in the measurement unit 14.

[0028] The control unit 120 includes a control circuit for controlling the operation of the inverter circuit 111, a memory for temporarily storing measurement data output from the measurement unit 114, and the like. The control circuit of the control unit 120 generates, for example, a PWM (Pulse Width Modulation) signal as a control signal for controlling the operation of the inverter circuit 111. The control unit 120 supplies the generated PWM signal to the gate terminal of the power semiconductor element of the inverter circuit 111, and controls the operation of the inverter circuit 111 by turning the power semiconductor element on / off.

[0029] The memory of the control unit 120 stores the measurement data of induced current and induced voltage output by the measurement unit 114. Preferably, the measurement data consists of a set of values ​​for induced current and induced voltage measured simultaneously by the measurement unit 114. The control unit 120 outputs the measurement data collected by storing it in memory to the communication unit 130. The measurement data collection period and the timing of output to the communication unit 130 are set as appropriate. In one example, the control unit 120 collects measurement data on a daily basis and outputs the collected measurement data to the communication unit 130 once a day.

[0030] The communication unit 130 is equipped with a communication interface for communicating with the remote diagnostic device 200 via a communication network NW. When the communication unit 130 acquires data to be transmitted from the control unit 120, it transmits the acquired data to the desired destination via the communication network NW. When it receives data transmitted from an external device (such as the remote diagnostic device 200 or user terminal 300) via the communication network NW, it outputs the received data to the control unit 120. In this embodiment, the communication unit 130 transmits the measurement data of induced current and induced voltage collected by the control unit 120 to the remote diagnostic device 200 via the communication network NW.

[0031] The remote diagnostic device 200 acquires measurement data of induced current and induced voltage from the power supply equipment 100 via a communication network NW. Based on the acquired measurement data of induced current and induced voltage, the remote diagnostic device 200 diagnoses the power supply equipment 100. More specifically, the remote diagnostic device 200 detects anomalies in the reactor 112 connected to the power conversion circuit 110 of the power supply equipment 100. Here, anomalies refer to abnormal conditions, including at least deterioration or lifespan, and not only mean a state that is not normal (usual), but may also include a state that is currently within the normal range but is moving towards or about to move towards an abnormal (malfunction) state and is therefore "anomaly". In other words, anomaly detection may also include detecting (judging) signs that are moving towards an abnormal situation.

[0032] In this embodiment, the power supply equipment 100 is configured to include an inverter circuit and an LC filter circuit (reactor 112 and capacitor 113). Alternatively, the power supply equipment 100 may be configured to include a DC / DC converter and a DC / AC inverter, or a PWM converter and a PWM inverter. The power supply equipment 100 is not limited to stationary power supply equipment, but may also include power supply equipment such as power conditioners installed in electric vehicles, energy storage systems (ESS), or V2X systems (Vehicle to Load, Home, Grid) including power conditioners for electric vehicles.

[0033] In this embodiment, the power supply unit 100 is configured to include a control unit 120 and a communication unit 130. Alternatively, the control unit 120 and the communication unit 130 may be provided outside the power supply unit 100.

[0034] In this embodiment, the control unit 120 is configured to collect measurement data. Alternatively, the control unit 120 may be configured to sequentially upload the measurement data to an external server (not shown) and have the external server collect the measurement data. The external server can then transmit the collected measurement data to the remote diagnostic device 200 at an appropriate time (for example, once a day).

[0035] Figure 2 is a block diagram showing the internal configuration of the remote diagnostic device 200. The remote diagnostic device 200 comprises a control unit 201, a storage unit 202, a communication unit 203, an operation unit 204, and a display unit 205.

[0036] The control unit 201 consists of a CPU (Central Processing Unit), ROM (Read Only Memory), RAM (Random Access Memory), and the like. The CPU in the control unit 201 loads various computer programs stored in the ROM or memory unit 202 onto the RAM and executes them, thereby enabling the entire device to function as a remote diagnostic device 200.

[0037] The control unit 201 is not limited to the above configuration and may be any processing circuit or arithmetic circuit equipped with multiple CPUs, multi-core CPUs, GPUs (Graphics Processing Units), microcontrollers, volatile or non-volatile memory, etc. The control unit 201 may also be equipped with functions such as a timer for measuring the elapsed time from the time a measurement start instruction is given until a measurement end instruction is given, a counter for counting numbers, and a clock for outputting date and time information.

[0038] The storage unit 202 is equipped with a storage device such as an HDD (Hard Disk Drive) or SSD (Solid State Drive). The storage unit 202 stores various computer programs executed by the control unit 201, as well as data necessary for the execution of these computer programs. The computer program stored in the storage unit 202 is, for example, an anomaly detection program PG that acquires measurement data of induced current and induced voltage induced by a reactor 112 connected to the power conversion circuit 110 of the power supply device 100 via a communication network NW, calculates the amount of change in impedance in the reactor 112 over time based on the acquired measurement data, and causes the computer to execute a process to detect anomalies in the reactor based on the calculated amount of change in impedance.

[0039] The computer program stored in the memory unit 202 may be provided on a non-temporary recording medium RM on which the computer program is recorded in a readable format. The recording medium RM is a portable memory such as a CD-ROM, USB (Universal Serial Bus) memory, SD (Secure Digital) card, microSD card, or CompactFlash®. In this case, the control unit 201 reads the computer program from the recording medium RM using a reading device (not shown) and installs the read computer program into the memory unit 202. Alternatively, the computer program may be provided via communication. In this case, the control unit 201 can download the computer program via the communication unit 203 and install the downloaded computer program into the memory unit 202.

[0040] The communication unit 203 is equipped with a communication interface for communicating with the communication unit 130 of the power supply device 100 via the communication network NW. When the communication unit 203 receives data (measurement data of induced current and induced voltage) transmitted from the communication unit 130 of the power supply device 100 via the communication network NW, it outputs the received data to the control unit 201. The control unit 201 stores the received data in the storage unit 202. When the control unit 201 inputs data to be sent to the communication unit 130 (for example, an instruction to the power supply device 100), the communication unit 203 transmits the input data to the power supply device 100 via the communication network NW.

[0041] The operation unit 204 is equipped with an input interface such as a keyboard and mouse, and accepts operations from the administrator of the power supply equipment 100. The display unit 205 is equipped with a liquid crystal display device or the like, and displays information that should be notified to the administrator of the power supply equipment 100. In this embodiment, the remote diagnostic device 200 is configured to include the operation unit 204 and the display unit 205, but the operation unit 204 and the display unit 205 are not essential, and the device may be configured to accept user operations via an external computer and transmit information that should be notified to the user to the external computer. An example of an external computer is the user terminal 300.

[0042] The following describes the anomaly detection process performed by the remote diagnostic device 200. (1) Acquisition of measurement data The control unit 120 of the power supply device 100 measures the induced current and induced voltage of the reactor 112 connected to the power conversion circuit 110 while the power supply device 100 is operating in the actual environment, and acquires the measurement data. The measurement data is time-series data. Hereinafter, the measurement data of the induced current measured while the power supply device 100 is operating in the actual environment will be denoted as i1(t), and the measurement data of the induced voltage will be denoted as v1(t).

[0043] The remote diagnostic device 200 acquires induced current measurement data i1(t) and induced voltage measurement data v1(t) from the power supply device 100 via the communication network NW. The acquired measurement data i1(t) and v1(t) are stored in the storage unit 202.

[0044] The memory unit 202 stores, in addition to the measurement data i1(t) and v1(t), the measurement data of the induced current and induced voltage of the reactor 112 measured in its initial state. The measurement data of the initial state will be denoted as i0(t) and v0(t), respectively. The initial state of the reactor 112 refers to, for example, the state in which the reactor 112 was first installed in the power supply equipment 100 (a state without deterioration over time or damage). The remote diagnostic device 200 acquires the measurement data i0(t) and v0(t) of the induced current and induced voltage measured for the reactor 112 as it was initially installed in the power supply equipment 100 from the power supply equipment 100, and stores the acquired measurement data i0(t) and v0(t) in the memory unit 202. Alternatively, the state of the reactor 112 at any point in time after the operation of the power supply equipment 100 has started may be used as the initial state.

[0045] (2) Calculation of impedance change The control unit 201 of the remote diagnostic device 200 calculates the change in impedance in the reactor 112 based on the measured data of induced current i0(t) and i1(t) and the measured data of induced voltage v0(t) and v1(t).

[0046] The control unit 201 performs Fourier analysis on the measured induced current data i0(t) and i1(t) and calculates the difference in the real parts of the obtained induced current frequency spectra. The difference in the real parts Re[Δi] is calculated as Re[Δi]=Re[i1(ω)]-Re[i0(ω)]. Re[i1(ω)] represents the real part of the frequency spectrum i1(ω) obtained by Fourier transforming i1(t), and Re[i0(ω)] represents the real part of the frequency spectrum i0(ω) obtained by Fourier transforming i0(t). Re[Δi] is obtained by taking the difference between the two. The Fourier transforms of i0(t) and i1(t) are performed on a computer using the Fast Fourier Transform (FFT) algorithm. Alternatively, algorithms using the Discrete Fourier Transform (DFT), Discrete Cosine Transform (DCT), etc., may be used.

[0047] Similarly, the control unit 201 performs Fourier analysis on the measured induced voltage data v0(t) and v1(t) and calculates the difference in the real parts of the frequency spectra of the obtained induced voltages. The difference in the real parts, Re[Δv], is calculated as Re[Δv]=Re[v1(ω)]-Re[v0(ω)]. Re[v1(ω)] represents the real part of the frequency spectrum v1(ω) obtained by Fourier transforming v1(t), and Re[v0(ω)] represents the real part of the frequency spectrum v0(ω) obtained by Fourier transforming v0(t). Re[Δv] is obtained by taking the difference between the two. The Fourier transforms of v0(t) and v1(t) are performed on a computer using the Fast Fourier Transform (FFT) algorithm. Alternatively, algorithms using the Discrete Fourier Transform (DFT), Discrete Cosine Transform (DCT), etc., may be used.

[0048] The control unit 201 calculates the impedance change amount Δz (=Re[Δv] / Re[Δi]) by dividing the above Re[Δv] by Re[Δi].

[0049] In this embodiment, a measuring coil (differential coil) is used to measure the induced current and induced voltage in the reactor 112. Ideally (even considering variations), the difference in induced current and the difference in induced voltage are close to zero, making it difficult to detect anomalies in the reactor 112 from the difference in induced current or induced voltage. In this embodiment, since Fourier transform is utilized, it is possible to separate the frequency components caused by the voltage and current of the reactor 112 from other frequency components. The voltage and current of the reactor contain high-frequency components, and these high-frequency components reflect the operating frequency of the power conversion circuit 110 (such as the switching frequency and carrier frequency). Therefore, the frequency components that are the focus when detecting anomalies are fixed to some extent. On the other hand, the other frequency components can be diagnosed as frequency components measured due to the influence of EMI (Electromagnetic Interference) emitted by the power conversion circuit 110, and they occupy a frequency band of kHz to GHz. In this embodiment, by utilizing the Fourier transform and focusing on the frequency components of the operating frequency (or the vicinity of the frequency components, or multiples of the frequency components), it becomes possible to detect anomalies in the reactor 112, and by tracking the time series of the amount of change, abnormalities in the reactor 112 are detected.

[0050] (3) Anomaly detection The control unit 201 detects an abnormality in the reactor 112 based on the calculated impedance change amount Δz. Specifically, the control unit 201 sets a threshold TH for the impedance change amount Δz and detects an abnormality in the reactor 112 by comparing the magnitude of the impedance change amount Δz with the threshold TH. For example, if the impedance change amount Δz is greater than the threshold TH, the control unit 201 can determine that an abnormality has occurred in the reactor 112 and prompt replacement or other action.

[0051] The threshold TH for the impedance change Δz can be set appropriately. For example, the threshold TH may be set by conducting experiments on the power supply equipment 100 in advance to obtain the impedance change Δz over time and identifying the impedance change Δz immediately before the power supply equipment 100 failed due to the degradation of the reactor 112. Alternatively, a first power supply circuit using a reactor that has not deteriorated significantly in the power conversion circuit and a second power supply circuit using a reactor that has deteriorated significantly in the power conversion circuit may be prepared, and the threshold TH may be set to classify the impedance change calculated based on the measured data of induced current and induced voltage obtained from the first power supply circuit and the impedance change calculated based on the measured data of induced current and induced voltage obtained from the second power supply circuit. When setting the threshold TH for classifying the impedance change, existing learning models such as logistic regression, k-nearest neighbors, decision trees, and support vector machines may be used, or statistical methods may be used.

[0052] In this embodiment, a threshold value TH is set for the impedance change amount Δz, and an abnormality in the reactor 112 is detected based on the relationship between the impedance change amount Δz and the threshold value TH. Alternatively, the threshold value TH may be set as a function of the elapsed time since the reactor 112 was installed in the power supply equipment 100, or as a function of the ambient temperature in the installation environment of the power supply equipment 100.

[0053] The operation of the remote diagnostic device 200 will be described below. Figure 3 is a flowchart illustrating the procedure performed by the remote diagnostic device 200. The control unit 201 of the remote diagnostic device 200 reads and executes the abnormality detection program PG from the storage unit 202, thereby performing the following processing. The storage unit 202 of the remote diagnostic device 200 is assumed to store measurement data i0(t) and v0(t) of the induced current and induced voltage of the reactor 112 measured in the initial state.

[0054] The control unit 201 of the remote diagnostic device 200 acquires measurement data i1(t) of the induced current induced by the reactor 112 and measurement data v1(t) of the induced voltage while the power supply device 100 is operating in the actual environment (step S101). For example, the control unit 201 can send a request to the power supply device 100 to transmit measurement data at an appropriate timing and receive the measurement data i1(t) and v1(t) transmitted from the power supply device 100 in response. Alternatively, the control unit 201 may receive the measurement data i1(t) and v1(t) that are spontaneously transmitted from the communication unit 130 of the power supply device 100. The control unit 201 stores the received measurement data i1(t) of the induced current and measurement data v1(t) of the induced voltage in the storage unit 202.

[0055] The control unit 201 calculates the change in impedance Δz in the reactor 112 based on the acquired measurement data i1(t) and v1(t) (step S102). The control unit 201 may calculate the change in impedance Δz each time measurement data i1(t) and v1(t) are acquired, or it may collect measurement data i1(t) and v1(t) for a predetermined period and then calculate the change in impedance Δz at predetermined intervals. The control unit 201 calculates the change in impedance Δz from the initial state using the initial state measurement data i0(t) and v0(t) that are stored in advance in the storage unit 202. Alternatively, the control unit 201 may calculate the change in impedance Δz from the previously calculated value.

[0056] The control unit 201 detects an abnormality in the reactor 112 based on the calculated impedance change Δz (step S103). The control unit 201 compares the impedance change Δz with the threshold TH, and if the impedance change Δz exceeds the threshold TH, it determines that an abnormality has occurred in the reactor 112. On the other hand, if the impedance change Δz does not exceed the threshold TH, the control unit 201 determines that there is no abnormality in the reactor 112.

[0057] The control unit 201 outputs information based on the detection result (step S104). If it is determined in step S103 that an abnormality has occurred in the reactor 112, the control unit 201 displays text information on the display unit 205 prompting the replacement of the reactor 112. Here, the text information may include symbolic information that can be recognized as prompting replacement, such as an error code. Figure 4 is a schematic diagram showing an example of the output of the detection result. Figure 4 shows an example in which text information indicating that an abnormality has been detected in the reactor 112 and that the replacement of the reactor 112 is recommended is displayed on the display unit 205. If it is determined that no abnormality has occurred in the reactor 112, the control unit 201 may display text information indicating that no abnormality has been detected in the reactor 112 on the display unit 205, or it may omit the processing in step S104.

[0058] If the remote diagnostic device 200 is equipped with an audio output means, the control unit 201 may prompt the replacement of the reactor 112 by outputting audio through the audio output means. Alternatively, the control unit 201 may notify the user terminal 300 via the communication unit 203 of information prompting the replacement of the reactor 112.

[0059] Furthermore, the control unit 201 may accept the user's decision on whether or not to accept the detected abnormality, and may present the user with a maintenance replacement time for the reactor according to the user's decision. For example, if the user accepts the abnormality of reactor 112, the control unit 201 may present an extension of the maintenance replacement time. If the user does not accept the abnormality of reactor 112, the control unit 201 may present an advancement of the maintenance replacement time.

[0060] As described above, the remote diagnostic device 200 according to Embodiment 1 detects abnormalities in the reactor 112 by acquiring measurement data of induced current and induced voltage induced by the reactor 112 from the power supply equipment 100 that is actually in operation, and by calculating the impedance change. If the remote diagnostic device 200 determines that an abnormality has occurred in the reactor 112, it can, for example, prompt the user to replace the reactor 112 without informing them of the abnormality, which can be useful in the maintenance plan for the power supply equipment 100.

[0061] (Embodiment 2) Embodiment 2 describes a configuration in which anomalies in the reactor 112 are detected using a machine learning model. The overall configuration of the remote diagnostic system 1 and the internal configuration of the remote diagnostic device 200 are the same as in Embodiment 1, so their explanation will be omitted.

[0062] The remote diagnostic device 200 according to Embodiment 2, upon input of measurement data i1(t) and v1(t) of induced current and induced voltage induced by the reactor 112, detects an anomaly in the reactor 112 using a learning model MD that has been trained to estimate anomalies in the reactor 112.

[0063] Figure 5 is a schematic diagram showing an example configuration of a learning model MD. The learning model MD comprises, for example, an input layer LY1, hidden layers LY2a and LY2b, and an output layer LY3. In the example in Figure 5, the learning model MD is configured to have two hidden layers LY2a and LY2b. Alternatively, the number of hidden layers may be one or three or more. An example of a learning model MD is a DNN (Deep Neural Network). Alternatively, SVM, XGBoost (eXtreme Gradient Boosting), LightGBM (Light Gradient Boosting Machine), etc. may be used.

[0064] Each layer constituting the learning model MD comprises one or more nodes. The nodes in each layer are unidirectionally coupled with the nodes in the preceding and succeeding layers with desired weights and biases. Vector data having the same number of components as the number of nodes in the input layer LY1 is provided as input data for the learning model MD. The input data in Embodiment 2 are measurement data i1(t) and v1(t) of induced current and induced voltage induced by the reactor 112 of the power supply equipment 100 operating in a real environment.

[0065] The data given to each node in the input layer LY1 is passed to the first hidden layer LY2a. In this hidden layer LY2a, the output is calculated using an activation function that includes weight coefficients and biases, and the calculated value is passed to the next hidden layer LY2b, and so on, until the output of the output layer LY3 is obtained.

[0066] The output layer LY3 outputs the probability P (0-100%) that reactor 112 is estimated to be heterogeneous.

[0067] The learning model MD is generated by learning according to a predetermined learning algorithm and determining the internal parameters of the learning model MD, including weight coefficients and biases. Prior to learning, a dataset is prepared that includes measurement data of induced current and induced voltage measured when no abnormality occurs in the reactor 112, and measurement data of induced current and induced voltage measured when an abnormality occurs in the reactor 112. The remote diagnostic device 200 uses the prepared dataset as training data and generates the learning model MD by learning using algorithms such as backpropagation and determining the internal parameters (weight coefficients, biases, etc.).

[0068] In the learning phase before operation begins, the remote diagnostic device 200 trains the learning model MD and stores the trained learning model MD in the storage unit 202. Alternatively, the learning model MD may be trained on an external server, and the trained learning model MD may be stored in the storage unit 202.

[0069] During the operational phase after the start of operation, when the remote diagnostic device 200 acquires measurement data i1(t) and v1(t) of induced current and induced voltage from the power supply equipment 100 in operation, it inputs the acquired measurement data i1(t) and v1(t) into the learning model MD. The control unit 201 of the remote diagnostic device 200 performs calculations by the learning model MD and detects abnormalities in the reactor 112 by referring to the information output from its output layer LY3. For example, the control unit 201 compares the probability P output from the output layer LY3 with a pre-set threshold THP, and determines that an abnormality has occurred in the reactor 112 if the probability P is greater than or equal to the threshold THP. If the control unit 201 determines that no abnormality has occurred in the reactor 112 if the probability P is less than the threshold THP.

[0070] In this embodiment, the learning model MD is stored in the storage unit 202, and the control unit 201 of the remote diagnostic device 200 performs calculations using the learning model MD. Alternatively, the learning model MD may be installed on an external server, and the system may access the external server via the communication unit 203 to have the calculations performed by the learning model MD performed on the external server. In this case, the control unit 201 of the remote diagnostic device 200 simply transmits the measurement data acquired from the operating power supply equipment 100 to the external server and has the calculations performed by the learning model MD performed on the external server.

[0071] As described above, the remote diagnostic device 200 according to Embodiment 2 detects abnormalities in the reactor 112 using a learning model MD. In Embodiment 2, since measurement data is collected from various power supply devices 100 with varying degrees of reactor 112 degradation, a learning model MD that is tailored to the actual environment can be constructed, enabling accurate detection of abnormalities in the reactor 112.

[0072] (Embodiment 3) Embodiment 3 describes a configuration in which the remote diagnostic device 200 estimates the lifespan of the reactor 112. The overall configuration of the remote diagnostic system 1 and the internal configuration of the remote diagnostic device 200 are the same as in Embodiment 1, so their explanation will be omitted.

[0073] The remote diagnostic device 200 according to Embodiment 3 estimates the probability P that the reactor 112 is heterogeneous using the learning model MD described above at regular intervals (for example, every month), and estimates the lifetime of the reactor 112 by deriving the time-series change of probability P.

[0074] Figure 6 is a graph showing the change in probability P over time. In the graph shown in Figure 6, the horizontal axis represents elapsed time (number of months), and the vertical axis represents the probability P that reactor 112 is heterogeneous. The control unit 201 of the remote diagnostic device 200 estimates the probability P that reactor 112 is heterogeneous using the learning model MD at regular intervals (for example, every month) and plots it on the graph. The control unit 201 finds a curve showing the time-series change of probability P and estimates the number of months in which this curve intersects the lifetime estimation threshold (80% in the example of Figure 6) as the lifetime of reactor 112. The control unit 201 displays the estimated lifetime information on the display unit 205.

[0075] The control unit 201 may present the user with the maintenance replacement timing for the reactor 112 based on the estimated lifespan. In this case, the control unit 201 should determine the maintenance replacement timing for the reactor to be the timing corresponding to the estimated lifespan. The remote diagnostic device 200 can present the determined maintenance replacement timing information to the user by displaying it on the display unit 205 or by notifying the user terminal 300.

[0076] As described above, in Embodiment 3, the lifespan of the reactor 112 can be estimated, and the maintenance replacement timing can be presented to the user, which can be useful for maintenance planning.

[0077] In Embodiments 1 and 2, a configuration for detecting an abnormality in the reactor 112 was described in the remote diagnostic device 200, and in Embodiment 3, a configuration for estimating the lifespan of the reactor 112 was described. Alternatively, the control unit 120 of the power supply equipment 100 may detect the abnormality in the reactor 112, or the control unit 120 of the power supply equipment 100 may estimate the lifespan of the reactor 112. The control unit 120 acquires measurement data of the induced current and induced voltage induced by the reactor 112 with respect to the power supply equipment 100 operating in a real environment, and can detect an abnormality in the reactor 112 by calculating the change in impedance Δz, etc. The control unit 120 may notify the user terminal 300, etc. of the detected abnormality in the reactor 112 via the communication unit 130. Furthermore, the control unit 120 may estimate the lifespan of the reactor 112 and notify the user terminal 300, etc. of the estimated lifespan via the communication unit 130.

[0078] The disclosed embodiments are illustrative in all respects and not restrictive. The scope of the invention is defined by the claims and includes all modifications in the sense and scope equivalent to the claims. [Explanation of Symbols]

[0079] 100 Power equipment 110 Power Conversion Circuit 111 Inverter Circuit 112 Reactor 113 Capacitors 114 Measurement Unit 120 Control Unit 130 Communication Unit 200 Remote diagnostic devices 201 Control Unit 202 Storage section 203 Communications Department 204 Operation section 205 Display section PG Anomaly Detection Program RM recording media

Claims

1. An acquisition unit that acquires measurement data obtained by measuring the induced current and induced voltage induced by a reactor connected to the power conversion circuit of a power supply device in a time series, A calculation unit that calculates the amount of change in impedance over time in the reactor based on the acquired measurement data, A detection unit detects the anomaly of the reactor based on the calculated change in impedance. A detection device equipped with the following features.

2. A learning unit that learns the criteria for detecting abnormalities in the reactor. The detection device according to claim 1, comprising:

3. An estimation unit estimates the time-series change in the probability of an abnormality occurring in the reactor using the measurement data, and estimates the lifespan of the reactor based on the estimated time-series change. The detection device according to claim 1, comprising:

4. A receiving unit that receives the user's judgment as to whether or not to accept the abnormality detected by the detection unit, A display unit that presents to the user the maintenance replacement timing for the reactor according to the user's judgment result. The detection device according to claim 3, comprising:

5. Power conversion circuit, A reactor connected to the power conversion circuit, A measurement unit that measures the induced current and induced voltage induced by the reactor in a time series. Power supply equipment equipped with, An acquisition unit that acquires the measurement data of the induced current and the induced voltage measured by the measurement unit from the power supply equipment via a communication network, A calculation unit that calculates the amount of change in impedance over time in the reactor based on the acquired measurement data, A detection unit detects the anomaly of the reactor based on the calculated change in impedance. Detection device equipped with A detection system that includes this.

6. By measuring the induced current and induced voltage from a reactor connected to the power conversion circuit of a power supply device in a time series, measurement data is obtained. Based on the acquired measurement data, the amount of change in impedance over time in the reactor is calculated. Based on the calculated change in impedance, the anomaly of the reactor is detected. A computer program that causes a computer to perform a process.

Images