JP2020009184A - Abnormality detection method and abnormality detection system - Google Patents

Abstract

To provide an abnormality detection method and an abnormality detection system capable of highly precisely detecting an abnormality in an electrical facility without being affected by an external environment such as an environmental change in an installation site of the electrical facility.SOLUTION: An abnormality detection method includes: a first step of acquiring measurement data from each of sensors 110 to 116 disposed in electric facilities 120 to 128; a second step of determining a representative value from the acquired measurement data; a third step of subtracting the representative value from the measurement data so as to obtain difference data; a fourth step of using differential data obtained by repeating the first to third steps over a predetermined period of time, to calculate a movement average and movement standard deviation for each sensor; and a fifth step of determining presence or absence of an abnormality by comparing the calculated movement average and movement standard deviation with thresholds. Accordingly, an abnormality in an electrical facility can be highly precisely detected without being affected by an external environment.SELECTED DRAWING: Figure 1

Description

  The present invention detects an abnormality in electrical equipment such as a switchboard, and detects an abnormality such as a stoppage of operation of the electrical equipment, and an abnormality detection method and an abnormality detection system capable of preventing a serious accident such as a fire from occurring. About.

  Electrical equipment such as a switchboard may be installed in a poor environment such as being installed outdoors or in a place exposed to direct sunlight, and is used for a long time in the same environment after installation. When an abnormality occurs in an electrical facility such as a switchboard, all devices to which power is supplied are affected. Further, occurrence of an abnormality may lead to a serious accident such as a fire. Therefore, in order to detect the occurrence of an abnormality in the electrical equipment in advance, a monitoring device for monitoring the temperature or the like is provided.

  For example, in Patent Literature 1 below, an odor generating component in which a nonflammable odor generating substance is sealed and an odor sensor that detects an odor substance radiated from the odor generating substance when overheating occurs are installed in a place where overheating is assumed. A technique for detecting overheating is disclosed.

  Further, Patent Literature 2 below discloses a technology for appropriately supplementing and reporting the presence or absence of an abnormality with respect to an electric facility or its installation space. Specifically, in the equipment abnormality alarm method, a standard pattern consisting of an average value and a standard deviation value of each temperature data at fixed time intervals is measured from temperature data measured by an optical fiber type temperature measuring means and stored in a time series. Is created, and it is determined that the temperature is abnormal when the measured temperature value deviates from the standard pattern.

JP-A-2003-240649 JP-A-10-11681

  However, in Patent Literature 1, overheating can be detected only at the place where the odor generating component is installed, and there is a problem that a power outage operation is required every time the odor generating component is installed or replaced.

  Further, in electrical equipment and the like installed outdoors, the standard pattern itself largely fluctuates daily due to the external environment (sunshine, wind and rain, etc.). There is a problem that cannot be done.

  Therefore, the present invention provides an abnormality detection method and an abnormality detection system that can accurately detect an abnormality of an electric facility without being affected by an external environment such as an environmental change in a place where the electric facility (including electric equipment) is installed. The purpose is to provide.

  An abnormality detection method according to a first aspect of the present invention includes a first step of acquiring measurement data measured at the same timing from each of a plurality of sensors arranged in electrical equipment, and a plurality of acquired measurement data. A second step of determining a representative value from a plurality of measurement data, a third step of subtracting a representative value from each of the plurality of measurement data to calculate difference data, and repeatedly executing the first, second, and third steps for a predetermined period of time. A fourth step of calculating at least one of a moving average value and a moving standard deviation value for each sensor using the difference data obtained by performing the calculation, and at least one of the calculated moving average value and the moving standard deviation value. , A fifth step of determining whether or not an abnormality has occurred in the electrical equipment by comparing with a predetermined threshold value.

  As described above, by using the difference between the measured value of each sensor and the representative value, the influence of the external environment (wind, rain, sunshine, etc.) can be canceled or reduced. By evaluating the presence or absence of an abnormality based on the moving average value and the moving standard deviation value, a short-term temporal change or a periodic temporal change (for example, a daily time change or a weekly time change) can be canceled or reduced. can do. Therefore, it is possible to accurately detect the abnormality of the electric equipment without being affected by the external environment.

  Preferably, in the fourth step, a moving average value and a moving standard deviation value are calculated, and in the abnormality detection method, a divergence value of difference data is calculated from the moving average value and the moving standard deviation value; Determining the presence or absence of an abnormality in the electrical equipment by comparing the absolute value of the current value with a predetermined threshold value, wherein L is a predetermined constant, and the difference between the i-th sensor among the plurality of sensors is determined. The data is Bi, the moving average value calculated using the difference data within a predetermined period including the difference data calculated before the difference data Bi as the latest difference data is set as AVBim, and the moving average value is set before the difference data Bi. The deviation of the difference data Bi is defined as σBim, where the moving standard deviation calculated using the difference data within a predetermined period including the calculated difference data as the latest difference data is σBim. Qi is Qi = 0 if (AVBim−L · σBim) ≦ Bi ≦ (AVBim + L · σBim), and Qi = Bi− (AVBim−L · if AVBim−L · σBim)> Bi. σBim), and if Bi> (AVBim + L · σBim), then Qi = Bi− (AVBim + L · σBim).

  As a result, it is possible to detect the possibility of an abnormality in the electrical equipment accompanied by a sudden change.

  More preferably, the predetermined period is a period that is a natural number multiple of 7 days.

  This makes it possible to cancel a weekly trend included in the difference data with respect to the electrical equipment such as a factory that is closed on Saturday and Sunday, so that the abnormality of the electrical equipment can be detected with higher accuracy. .

  More preferably, the sensor is a temperature sensor, and the height of the arrangement position of the plurality of sensors on the electric equipment is the same as each other.

  Thus, the influence of the external environment (such as the influence of sunlight) on each temperature sensor becomes the same, so that it is easy to detect the abnormality of the electric equipment.

  Preferably, the electrical equipment is composed of a plurality of electrical devices, and among the measurement data of the temperature sensor, the representative value is determined by excluding the measurement data of the temperature sensor arranged in the electrical device having the temperature control device. .

  Thereby, the representative value can be determined more appropriately, and the detection accuracy of the abnormality can be improved.

  An abnormality detection system according to a second aspect of the present invention includes a plurality of sensors arranged in electrical equipment, a data collection device that acquires measurement data from each of the plurality of sensors at the same timing for a predetermined period, and a data collection unit. An analysis device for analyzing the acquired measurement data. The analyzer determines a representative value from a plurality of measurement data measured at the same timing, subtracts a representative value from each of the plurality of measurement data measured at the same timing, calculates difference data, and obtains the difference data during a predetermined period. Using the calculated difference data, at least one of a moving average value and a moving standard deviation value is calculated for each sensor, and at least one of the calculated moving average value and the moving standard deviation value is compared with a predetermined threshold value. By doing so, it is determined whether or not an abnormality has occurred in the electrical equipment.

  Thereby, the abnormality of the electric equipment can be accurately detected without being affected by the external environment.

  Each of the plurality of sensors has a wireless communication function for transmitting measurement data to the data collection device.

  Thereby, the temperature sensor can be easily installed in the existing electric equipment (without power failure depending on conditions).

  ADVANTAGE OF THE INVENTION According to this invention, the abnormality of an electrical equipment can be detected accurately, without being influenced by an external environment. Further, by using the divergence value, it is possible to detect the possibility of an abnormality (such as overheating) of the electric equipment accompanied by a rapid change.

It is a block diagram showing a schematic structure of an abnormality detection system concerning an embodiment of the invention. FIG. 2 is a block diagram illustrating an internal configuration of the temperature sensor of FIG. 1. FIG. 2 is a block diagram illustrating an internal configuration of a data collection device and an analysis device of FIG. 6 is a flowchart illustrating a process executed by the analysis device. 6 is a graph for explaining a method of calculating a moving average value and a moving standard deviation value. It is a block diagram showing arrangement of a temperature sensor concerning a modification. It is a graph which shows the measurement data of the 1st-3rd sensor among a total of five temperature sensors arrange | positioned at indoor equipment, and its analysis result. It is a graph which shows the measurement data of the 4th and 5th sensors among a total of 5 temperature sensors arrange | positioned at indoor equipment, and the analysis result. 8 is a graph showing difference data calculated from the measurement data of the first sensor shown in FIG. 7 and a simulation result in which a temperature abnormality has occurred in a part of the measurement data of the first sensor. It is a graph which shows the measurement data of the 1st-4th sensor among the total 7 temperature sensors arrange | positioned in outdoor facilities, and the analysis result. It is a graph which shows the measurement data of the 5th sensor among the total seven temperature sensors arrange | positioned in outdoor facilities, and the analysis result. 11 is a graph showing difference data calculated from the measurement data of the second sensor shown in FIG. 10 and a simulation result in which a temperature abnormality has occurred in a part of the measurement data of the second sensor.

  In the following embodiments, the same components are denoted by the same reference numerals. Their names and functions are also the same. Therefore, detailed description thereof will not be repeated.

(Configuration of abnormality detection system)
With reference to FIG. 1, an abnormality detection system 100 according to an embodiment of the present invention includes a first temperature sensor 110 to a fifth temperature sensor 118 arranged on each of a first switchboard 120 to a fifth switchboard 128, It includes a collection device 130 and an analysis device 132. The first to fifth switchboards 120 to 128 are switchboards of the same type or similar switchboards, and are a plurality of switchboards connected and arranged (hereinafter, also referred to as row boards). The first temperature sensor 110 to the fifth temperature sensor 118 are arranged at substantially the same position (at substantially the same height) on the upper surface of the first switchboard 120 to the fifth switchboard 128.

  Referring to FIG. 2, first temperature sensor 110 includes a control unit 140, a storage unit 142, a communication unit 144, a timer 146, a bus 148, a sensor element 150, and an A / D conversion unit 152. The control unit 140 controls each unit constituting the first temperature sensor 110, and is, for example, a CPU (Central Processing Unit), a microcomputer (hereinafter, referred to as a microcomputer), or the like. The storage unit 142 stores data and is, for example, a rewritable nonvolatile semiconductor memory. The communication unit 144 is a wireless communication module for communicating with the outside. In response to a request from the control unit 140, the timer 146 transmits information indicating the current time (hereinafter, simply referred to as the current time) to the control unit 140. The sensor element 150 is an element for measuring temperature, and outputs an analog signal corresponding to the temperature detected by the sensor element 150. The sensor element 150 is, for example, a resistance temperature detector, a thermocouple, or the like. The A / D converter 152 converts an analog signal output from the sensor element 150 into digital data at a predetermined timing and outputs the digital data. Digital data (temperature measurement data) output from the A / D conversion unit 152 is stored in the storage unit 142 in time series. Data exchange between the units is performed via a bus 148. Power is supplied to each unit by incorporating a battery or by a service outlet provided in each switchboard.

  The second to fifth temperature sensors 112 to 118 have the same configuration as the first temperature sensor 110. The communication units of the first to fifth temperature sensors 110 to 118 have information (communication addresses and the like) for uniquely distinguishing each of them. In addition, the time of the timer of each of the first temperature sensor 110 to the fifth temperature sensor 118 is assumed to be the same as the time of the timer of the analyzer 132 (to be described later). The same time (coordinated time) means not only that they are completely the same time, but also includes a case where they coincide within a predetermined allowable range.

  Referring to FIG. 3, the configurations of data collection device 130 and analysis device 132 are shown. The data collection device 130 includes a control unit 160, a storage unit 162, a communication unit 164, and a bus 166. The control unit 160 controls each unit configuring the data collection device 130, and is, for example, a CPU, a microcomputer, or the like. The storage unit 162 stores data and is, for example, a rewritable nonvolatile semiconductor memory. The communication unit 164 is a communication module for communicating with the first to fifth temperature sensors 110 to 118 and the analyzer 132. The communication unit 164 has a wireless communication function for communicating with the first to fifth temperature sensors 110 to 118. The data collection device 130 has a wired communication function (for example, LAN, USB, GPIB, RS-232C, etc.) for communicating with the analysis device 132. Data exchange between the units constituting the data collection device 130 is performed via a bus 166.

  The analysis device 132 includes a control unit 170, a storage unit 172, a communication unit 174, a timer 176, a display unit 178, an operation unit 180, and a bus 182. The control unit 170 controls each unit constituting the analysis device 132, and is, for example, a CPU. The storage unit 172 stores data and is, for example, a rewritable nonvolatile semiconductor memory. The communication unit 174 is a wired communication module for communicating with the data collection device 130. The timer 176 transmits the current time to the control unit 170 in response to a request from the control unit 170. The display unit 178 displays visual information (text, image, etc.). The display unit 178 includes, for example, a display panel such as a liquid crystal display panel, and a driving circuit that drives each pixel of the display panel. The operation unit 180 is for inputting an instruction to the analysis device 132, and is, for example, a keyboard. Data exchange between the units constituting the analyzer 132 is performed via a bus 182. The analysis device 132 may be, for example, a computer. The data collection 130 or the data collection device 130 and the analysis device 132 may be, for example, a sequencer.

(Operation of the abnormality detection system)
Hereinafter, the first temperature sensor 110 will be described as a representative of the plurality of temperature sensors. The second to fifth temperature sensors 112 to 118 operate similarly to the first temperature sensor 110.
(Operation of temperature sensor)

  The first temperature sensor 110 transmits temperature data (hereinafter, also referred to as measurement data) measured by the sensor element 150 at regular time intervals to the data collection device 130. This function of the first temperature sensor 110 is realized by the control unit 140 reading and executing a predetermined program stored in the storage unit 142. Specifically, the control unit 140 acquires the current time from the timer 146, determines whether or not the time for measuring the temperature (hereinafter, also referred to as measurement timing) has been reached. The A / D converter 152 is controlled to convert an analog signal (measurement signal of the sensor element 150) input to the A / D converter 152 into digital data, and store the digital data in the storage unit 142. Subsequently, the control unit 140 reads the measurement data from the storage unit 142 and transmits the measurement data to the data collection device 130 via the communication unit 144.

  For example, when the first temperature sensor 110 measures and transmits the surface temperature of the first switchboard 120 every one hour, the measurement start time and the measurement interval (for example, 60 minutes) are used as the measurement timing information. May be stored in the storage unit 142 in advance. The control unit 140 can read the information of the measurement timing from the storage unit 142 and refer to the current time by the timer 146 to realize the above processing.

Here, it is assumed that the same measurement timing is set for the first to fifth temperature sensors 110 to 118. As described above, since the timers of the first temperature sensor 110 to the fifth temperature sensor 118 have the same time, the first temperature sensor 110 to the fifth temperature sensor 118 respectively output the first switchboard 120 to the same at the same timing. The measurement data of the surface temperature of the fifth switchboard 128 is transmitted. Note that “same timing” means not only that the timings are completely the same, but also that the timings coincide within a predetermined allowable range, as is clear from the above description regarding the “same timing”. It is a meaning that includes.
(Operation of data collection device)

  The data collection device 130 receives the measurement data transmitted from the first temperature sensor 110 to the fifth temperature sensor 118 and transmits them to the analysis device 132. This function of the data collection device 130 is realized by the control unit 160 reading and executing a predetermined program stored in the storage unit 162. Specifically, control unit 160 controls communication unit 164 to repeatedly determine whether measurement data transmitted from first temperature sensor 110 to fifth temperature sensor 118 has been received. When measurement data is received from any of the first to fifth temperature sensors 110 to 118, the received measurement data is stored in the storage unit 162 so that the transmission source temperature sensor can be specified. For example, the received measurement data and the transmission address of the transmission source are stored in association with each other. Further, as described above, the measurement is performed from the first temperature sensor 110 to the fifth temperature sensor 118 at the same time, and the measurement data is transmitted. Therefore, the control unit 160 determines the measurement data measured at the same time. Are stored in the storage unit 162 in association with each other.

The data collection device 130 repeats the above processing until measurement data is received from all of the first to fifth temperature sensors 110 to 118. When receiving the measurement data from all of the first to fifth temperature sensors 110 to 118, the data collection device 130 reads the received measurement data from the storage unit 162 and transmits the measurement data to the analysis device 132 via the communication unit 164. . At this time, the data collection device 130 transmits the measurement data to the analysis device 132 in a format that makes it easy to know which temperature sensor is the data measured by which temperature sensor. For example, the measurement data is transmitted in association with information (transmission address, ID, etc.) for specifying the transmission source temperature sensor. Since the number of measurement data is fixed, the order of transmission or, when one packet includes a plurality of measurement data, the order in the packet is used as information for specifying the source temperature sensor. You may. For example, the corresponding measurement data may be transmitted in the order of the first to fifth temperature sensors 110 to 118.
(Operation of the analyzer)

  The analysis device 132 receives the measurement data of the first temperature sensor 110 to the fifth temperature sensor 118 from the data collection device 130, and divides the received measurement data in time series for each of the first temperature sensor 110 to the fifth temperature sensor 118. Is stored in the storage unit 172 (received in the past and added to the stored measurement data). The analysis device 132 executes the analysis process when the storage unit 172 stores as many measurement data items as can perform the analysis process described below for each of the first temperature sensor 110 to the fifth temperature sensor 118. The analysis device 132 determines whether any of the first to fifth switchboards 120 to 128 has an abnormality, or whether there is a sign of an abnormality, based on the analysis result.

  The analysis device 132 executes, for example, the processing illustrated in FIG. 4 are realized by the control unit 170 reading a predetermined program from the storage unit 172 and executing it.

  In step 300, the control unit 170 determines whether the communication unit 174 has received the measurement data from the analysis device 132. If it is determined that it has been received, control proceeds to step 302. Otherwise, the process of step 300 is repeated.

  In step 302, the control unit 170 stores the received measurement data in the storage unit 172. As described above, since the plurality of measurement data are transmitted from the analysis device 132 so as to know which one of the first temperature sensor 110 to the fifth temperature sensor 118 corresponds to each of the plurality of measurement data, the control unit 170 With respect to each of the temperature sensors 110 to the fifth temperature sensor 118, measurement data can be stored in the storage unit 172 in a time series.

  In step 304, the control unit 170 determines whether to execute the analysis. This is for determining the timing of executing the analysis first. Step 304 is repeatedly performed each time new measurement data is received. If it is determined that the analysis process is to be performed once, then if step 304 is subsequently performed, it is determined that the analysis is to be performed. The criteria for the determination can be set arbitrarily. For example, if a predetermined number or more of measurement data of each temperature sensor is stored in the storage unit 172, it is determined that the analysis is to be performed. Further, if a predetermined time has elapsed since the first reception of the measurement data, it may be determined that the analysis is to be performed. A predetermined flag is secured in a predetermined area of the storage unit 172, and if it is determined that the analysis is to be performed once, the flag is turned on (for example, a value different from the initial value is set in the flag), and thereafter, May determine that the analysis is always performed based on the value (ON) of the flag.

  In step 306, the control unit 170 reads, from the storage unit 172, a set of measurement data measured at the same timing with respect to each measurement data of the first temperature sensor 110 to the fifth temperature sensor 118, and stores a representative value thereof. The determined representative value is stored in the storage unit 172. Here, each measurement data of the first to fifth temperature sensors 110 to 118 is represented by Aij. i is an integer of 1 to 5, and represents the first to fifth temperature sensors 110 to 118, respectively. j is a positive integer representing the order of collection. Further, it is assumed that n pieces of measurement data are collected at equal time intervals during the predetermined period T for each of the first temperature sensor 110 to the fifth temperature sensor 118. The n pieces of measurement data in the predetermined period T are set as calculation targets of a moving average value and the like described later.

For example, for each measurement data Aij (i = 1 to 5, j = 1 to n) of the first temperature sensor 110 to the fifth temperature sensor 118, a representative value Rj is calculated by Rj = (Σ _i Aij) / 5. . Ｉ _i means an operator that takes the sum of Aij (j is the same value) for i = 1 to 5. That is, Rj (j = 1 to n) is an average value of five measurement data measured at the same timing. Note that the representative value is not limited to an average value using all five measurement data measured at the same timing. The average calculated using some of the measurement data may be used as the representative value.

  When step 306 is first executed, a representative value Rj (j = 1 to n) is calculated for each of n pieces of measurement data (5 × n in total) of the first to fifth temperature sensors 110 to 118. I do. After that, every time new measurement data is received in step 300, step 306 is executed. For the past measurement data, a representative value has already been calculated (stored in the storage unit 172). A representative value is calculated using the five measured data thus obtained, and stored in the storage unit 172. The same applies to the processing of steps 308 to 314 described later. When the first measurement is performed, the processing is performed for a total of 5 × n measurement data of the first temperature sensor 110 to the fifth temperature sensor 118, and thereafter, every time new measurement data is received in step 300, When executed, necessary processing is executed using the five newly received measurement data, and the already calculated result is read from the storage unit 172 and used.

  In step 308, the control unit 170 reads the measurement data of the first temperature sensor 110 to the fifth temperature sensor 118 for the predetermined period T from the storage unit 172, and calculates a difference by comparing the measurement data with the representative value. Then, the calculation result (difference data) is stored in the storage unit 172. Specifically, the corresponding representative value Rj is subtracted from the measurement data Aij to obtain a difference Bij (Bij = Aij-Rj (i = 1 to 5, j = 1 to n)). As a result, n pieces of difference data are stored in the storage unit 172 for each of the first to fifth temperature sensors 110 to 118.

  In step 310, the control unit 170 reads the difference data within the predetermined period T from the storage unit 172, calculates a moving average value AV for them, and stores the calculation result (moving average value) in the storage unit 172. Specifically, an average value of n pieces of difference data Bij (j = 1 to n) is calculated for the i-th temperature sensor. Step 310 is executed while changing target data every time new difference data is obtained in step 308. FIG. 5 shows difference data obtained from the measurement data of the i-th temperature sensor. The horizontal axis in FIG. 5 represents time, and the vertical axis represents difference data. The difference data is schematically indicated by a broken line. The actual difference data is digital data, and a part thereof is shown by a black circle in FIG. Regarding the difference data shown in FIG. 5, the moving average can be obtained by repeating the process of calculating the average value while shifting the target data for calculating the average value (data within the predetermined period T). For example, when the average value of n pieces of difference data during the time t1 to t3 (period T) is calculated, and then new difference data (difference data Bik at the time t4) is calculated, the time difference between the time t2 and the time t4 is calculated. The average value of the n pieces of difference data during the period (period T) is calculated. The predetermined period T is also called a window period.

  In step 312, as in step 310, the control unit 170 reads the difference data within the predetermined period T from the storage unit 172, calculates the moving standard deviation value σ for them, and stores the calculation result (moving standard deviation value). The information is stored in the unit 172. Specifically, a standard deviation value of n pieces of difference data Bij (j = 1 to n) is calculated for the i-th temperature sensor. The process of step 312 is repeatedly executed every time new difference data is obtained in step 308, as in step 310.

  In step 314, the control unit 170 determines whether to calculate a deviation value. As will be described later, the deviation value indicates the degree to which the current difference data deviates from the normal value, and is determined using the latest moving average value and the latest moving standard deviation value. “Nearest” means a value calculated using n consecutive difference data up to the immediately preceding difference data without including the difference data for which the divergence value is to be calculated. For example, with reference to FIG. 5, the divergence value of the difference data Bik is not the difference data between times t2 and t4, but the moving average value and the moving standard calculated using the difference data between times t1 and t3. It is determined using the deviation value. Therefore, when step 314 is executed for the first time, it is determined that the latest moving average value and the latest moving standard deviation value do not exist and the deviation value is not calculated, and the control proceeds to step 318. When step 314 is performed for the second time or later, it is determined that the deviation value is calculated, and the control proceeds to step 316.

In step 316, the control unit 170 updates the latest difference data Bi (Bi = Ai-R (R is a representative value of A1 to A5)) obtained from the latest measurement data Ai (i = 1 to 5), The moving average value AVBim (m is the number of times of calculation of the moving average value) and the latest moving standard deviation value σBim (m is the number of times of calculation of the moving standard deviation value) are read from the storage unit 172, and L is a predetermined constant. , The divergence value Qi of the difference data Bi is calculated, and the calculation result (divergence value) is stored in the storage unit 172.
If (AVBim−L · σBim) ≦ Bi ≦ (AVBim + L · σBim), then Qi = 0. Qi = 0 means that Bi is in the normal range.
If (AVBim−L · σBim)> Bi, then Qi = Bi− (AVBim−L · σBim).
If Bi> (AVBim + L · σBim), Qi = Bi− (AVBim + L · σBim).
Note that the value of L is arbitrary and can be set as appropriate, but it is preferable that L = 2 to 3.

  In step 318, the control unit 170 determines that at least one of the moving average value, the moving standard deviation value, and the deviation value calculated in each of steps 310, 312, and 316 is a predetermined value (hereinafter, referred to as a predetermined value) set for each of them. It is determined whether it is larger than the threshold value. The threshold value is, for example, a value obtained by adding a fixed value to an initial value or a normal value. If it is determined that at least one of the moving average value, the moving standard deviation value, and the deviation value (absolute value) is larger than the threshold value, the control proceeds to step 320. Otherwise, control proceeds to step 322. The threshold value set for each of the moving average value, the moving standard deviation value, and the deviation value determines the accuracy with which the occurrence of an abnormality is predicted (to minimize the possibility of erroneous detection). , Or more on safety).

  In step 320, the control unit 170 presents a message indicating an abnormality. For example, the control unit 170 displays an image including a predetermined message on the display unit 178. For example, if the moving average value exceeds the threshold value, a message indicating that an abnormal heat source exists (or possibly exists) in the corresponding switchboard is presented. If the moving standard deviation exceeds the threshold, a message indicating that (or possibly) an intermittent abnormal heat source is present in the corresponding switchboard is presented. If the deviation value exceeds the threshold value, a message indicating that an abnormality related to overheating has occurred (or is likely to occur) in the corresponding switchboard is presented.

  In step 322, control unit 170 determines whether an end instruction has been received. The end instruction is given, for example, by operating the operation unit 180. If it is determined that the termination instruction has been received, the program ends. Otherwise, control returns to step 300, and control section 170 repeats the above processing.

  As described above, the abnormality detection system 100 can accurately detect the temperature abnormality in the row board. That is, by using the average value of the temperature data of a plurality of switchboards as the representative value and using the difference between the measured value of each temperature sensor and the representative value, the influence of the external environment (wind, rain, sunshine, etc.) can be canceled or reduced. it can. Further, with the moving average value and the moving standard deviation value, a short-term temporal change or a periodic temporal change (for example, a temporal change every day or every week) of the difference can be canceled or reduced, and an abnormality occurs. Can be accurately determined.

  By checking the time change (hereinafter also referred to as trend) of each of the moving average value and the moving standard deviation value, when there is a certain amount of increase from the initial value or the normal value, it is determined that the power distribution panel is abnormal. It is possible to appropriately detect the presence or absence of a sign of abnormality due to the gradual deterioration of the electrical equipment. If there is no significant change in the trends of the moving average value and the moving standard deviation value, it can be determined that there is a high possibility that the soundness of the electrical equipment is secured.

  Further, by using the divergence value calculated as described above, it is possible to detect the possibility of an abnormality (such as overheating) of the electrical equipment accompanied by a rapid temperature change.

  As described above, by using the temperature sensor having the wireless communication function, the temperature sensor can be easily installed in the existing electrical equipment (without power interruption depending on conditions).

(Modification)
In the above, the case where one temperature sensor is arranged in each of a plurality of switchboards has been described, but the present invention is not limited to this. A plurality of temperature sensors may be arranged on each panel. Further, it is not necessary to dispose the temperature sensor on some of the switchboards constituting the row board. For example, there is a case where a plurality of switchboards, which can be considered to have almost the same condition, measure the temperature with one temperature sensor as a representative. In such a case, it is preferable to determine the representative value by weighting the measurement data of each temperature sensor.

  For example, FIG. 6 illustrates a configuration in which the sixth switchboard 202 and the seventh switchboard 204 are added to the configuration of the first switchboard 120 to the fifth switchboard 128 in which the first temperature sensor 110 to the fifth temperature sensor 118 illustrated in FIG. The configuration shown is shown. A sixth temperature sensor 200 is arranged on the sixth switchboard 202, but no temperature sensor is arranged on the seventh switchboard 204. In such a case, if the measurement data of each temperature sensor is Ai (i = 1 to 6), the representative value R may be calculated by R = (A1 + A2 + A3 + A4 + A5 + 2 · A6) / 7. The weight of the measurement data of the first temperature sensor 110 to the second temperature sensor 112 is “1”, and the weight of the measurement data of the sixth temperature sensor 200 arranged on the sixth switchboard 202 is “2”.

  The total number of temperature sensors arranged on the row board is arbitrary, but is preferably 5 or more. When the total number of the temperature sensors is small, the abnormality of the measurement data due to the temperature abnormality of a specific board greatly affects the representative value, and thus the difference data of other boards is easily affected. If the total number of temperature sensors is 5 or more, such an influence can be suppressed.

  In addition, a temperature control device such as a cooler may be arranged only on a specific board of the row boards. The temperature trend of such a particular board can be significantly different from the temperature trends of other boards. Therefore, it is preferable that measurement data of such a board is not used as data for determining a representative value.

  In the above description, a row board composed of five switchboards is shown, but the present invention is not limited to this. It may be composed of four or less or six or more switchboards. Elements constituting the train board are not limited to the switchboard. Electrical equipment other than the switchboard may be used.

  Further, the present invention is not limited to the side panel in which the same or similar electric facilities are arranged in parallel. It may be a plurality of electric facilities arranged in close proximity. Further, a plurality of temperature sensors may be arranged in one electric facility.

  In the above, the case where temperature abnormality is detected by the temperature sensor has been described, but the present invention is not limited to this. For example, instead of or in addition to the temperature sensor, a sensor other than the temperature sensor (for example, a humidity sensor, an acceleration sensor, etc.) is arranged on the train, and the moving average is obtained using the measurement data as described above. A value, a moving standard deviation value, and a deviation value may be calculated to detect an abnormality in the train. When the acceleration sensor is used, even if the output data (acceleration) of the acceleration sensor itself is to be analyzed, the vibration frequency calculated from the acceleration (for example, a ratio (%) of an occurrence period of an acceleration of a predetermined value or more in a certain period) May be the analysis target.

  In the above, the case where each temperature sensor is arranged on the upper surface of each switchboard and the height of the mounting position of each temperature sensor is substantially equal has been described, but the mounting position of each temperature sensor is arbitrary. By arranging the temperature sensor above the switchboard, it becomes easy to detect a temperature abnormality such as overheating in the switchboard. In addition, by arranging the temperature sensors at substantially the same height, the influence of the surrounding environment (sunshine or the like) on the temperature sensors can be made uniform, and it becomes easier to detect a temperature abnormality in the switchboard.

  In the above, the case where the moving average value, the moving standard deviation value, and the deviation value are all calculated has been described, but the present invention is not limited to this. An abnormality can be detected by calculating at least one of a moving average value, a moving standard deviation value, and a deviation value, and comparing the calculated value with a threshold value.

  In the above description, the calculation of the moving average value AVBim and the moving standard deviation value σBim in the analyzer 130 is performed every time the measurement data is received, but the calculation is not limited to this. After storing the measurement data for a certain period, the measurement may be performed collectively. For example, the measurement data may be performed every day, and the deviation value Qi of the next day is calculated based on the calculated moving average value AVBim and the calculated moving standard deviation value σBim. May be used. By doing so, the burden on the CPU and the like required for the calculation can be reduced.

  In the above, the case where the communication between the data collection device 130 and the analysis device 132 is a wired communication has been described, but the present invention is not limited to this. Communication between the data collection device 130 and the analysis device 132 may be wireless communication. Further, the analysis device 132 may have a wireless communication function, communicate directly with the first to fifth temperature sensors 110 to 118, and execute a data collection function of the data collection device 130.

  Communication between the temperature sensor and the data collection device may be wire communication. In that case, the data collection device may be mounted on the row board or arranged near the row board, and data may be transmitted from the data collection apparatus to the analysis apparatus by wireless communication. If the distance between the data collection device and the analysis device is long, in the case of a wired connection, a communication method capable of long-distance transmission such as RS422 and RS485 may be used.

  In the above, the case where the first temperature sensor 110 to the fifth temperature sensor 118 transmit the measurement data every time the temperature is measured has been described, but the present invention is not limited to this. The measurement data of a plurality of times may be stored and transmitted to the data collection device 130 collectively. Since the measurement timing of each of the first temperature sensor 110 to the fifth temperature sensor 118 is the same, when transmitting a plurality of data collectively, the measurement order between the plurality of transmitted data is determined by the data collection device 130. If it is understood from the above, the representative value can be determined using the measurement data of the first to fifth temperature sensors 110 to 118 measured at the same timing.

  In the above description, each temperature sensor has a timer and a storage unit, and sends the measurement data to the data collection device 130 at regular intervals. For example, the data collection device 130 has a timer and communicates with each temperature sensor at regular intervals (for example, every hour) (for example, by requesting each temperature sensor to transmit measurement data) to acquire the current value of the temperature. Alternatively, data logging (stored as time-series data) may be performed in the storage unit 162 of the data collection device 130.

  Further, in the above description, the communication between the data collection device 130 and the analysis device 132 is a direct communication system, but this is not essential. For example, the collected measurement data is stored in a predetermined data server or cloud via radio waves such as a mobile phone or the Internet, and the data in the data server or cloud is read out by the analyzer 132 to perform data analysis. You may.

  The case where the message is presented on the display unit 178 has been described above, but the present invention is not limited to this. It may be presented by sound (including voice) or lighting of an LED or the like. Further, the occurrence of an abnormality may be presented by display on a remote device such as a central monitoring panel or by sending an e-mail.

  The experimental results are shown below, and the effectiveness of the present invention is shown. In the same manner as in FIG. 1, a temperature sensor was arranged on each panel of a row panel composed of five electric facilities installed indoors, and the temperature was measured. The results of analyzing the measured temperature data as described above are shown in FIGS. 7 and 8, (1) to (5) shown in the column of “No.” at the left end are located at the temperature centers arranged similarly to the first temperature sensor 110 to the fifth temperature sensor 118 in FIG. 1. Corresponding. FIG. 9 shows the results of analysis and simulation using the measurement data of (1) in FIG.

  In FIG. 7, the graph in the column of “Temperatures A1 to A3” shows the actually measured values of the temperature sensor. The vertical axis is temperature, and the horizontal axis is days. The temperature was measured every hour for a period of 200 days or more of the sixth temperature sensor. The graph in the column of “moving average of difference AVB1 to AVB3” is the difference data obtained by using the respective measurement data of “temperatures A1 to A3” and calculating the average value of the five measurement data as a representative value as described above. Shows the moving average value calculated using. For calculating the moving average, the window period was set to two weeks (14 days), and 336 (= 14 × 24) consecutive data were used. The graphs in the column of “moving standard deviation of difference σB1 to σB3” show the moving standard deviation calculated using the same difference data (data for two weeks) for which the moving average has been calculated. The same applies to FIG.

  As can be seen from FIGS. 7 and 8, the actually measured temperature data has a change (seasonal change) of about 20 degrees for any temperature sensor, but the moving average value is almost the same for any temperature sensor. It has a constant value. As will be described later with reference to FIG. 9, in the difference data, a long-term temperature change due to the season is canceled, but a relatively large daily temperature change remains. On the other hand, in the moving average value, the daily temperature change is canceled or reduced. If there is no abnormal overheating in the panel, the moving average value is almost constant. Even in the moving standard deviation value, the daily temperature change is canceled or reduced. The moving standard deviation value is an index indicating the temperature variation that normally occurs with respect to the moving average value, and its change is relatively small unless there is abnormal overheating or the like in the panel. By observing the trends of the moving average value and the moving standard deviation value during normal times, it is possible to appropriately set the respective threshold values for abnormal heat generation.

  The vertical and horizontal axes of the graph shown in FIG. 9 are the same as those in FIG. The uppermost graph in FIG. 9 shows the difference calculated from the measurement data shown in the upper left graph of FIG. 7 (the measurement data of the temperature sensor in (1)) using the average value of the measurement data of the five temperature sensors as the representative value. Show data. As described above, in the difference data, a long-term temperature change due to the season is canceled, but a relatively large daily temperature change remains.

  The graphs at the second and third rows from the top in FIG. 9 are the same as the graphs at the center and right end at the top of FIG. The fourth and fifth graphs from the top in FIG. 9 show simulation results simulating the occurrence of abnormal heat generation. Specifically, it is assumed that abnormal heat generation occurs in each of the measurement data (A1) between the operation unit 180 days and the sixth temperature sensor 200 days (the period indicated by the arrow in FIG. 9) counted from the measurement start. 3 ° C was added. The results of calculating the moving average value and the moving standard deviation value using the data in the same manner as above are the graphs in the fourth and fifth rows from the top in FIG. 9, respectively. As can be seen from a comparison between the second graph and the fourth graph showing the moving table average value, a clear change occurs in the moving table average value in the period in which 3 ° C. is added. Similarly, as can be seen from a comparison between the third graph and the fifth graph from the top showing the moving standard deviation value, in the period in which 3 ° C. is added, a clear change occurs in the moving standard deviation value. I have. As described above, for a row board installed indoors, the occurrence of a temperature abnormality can be clearly detected from the moving average value and the moving standard deviation value calculated from the measurement data of the surface temperature of each board.

  A temperature sensor was arranged on each of the row boards installed outdoors, temperature was measured in the same manner as in Example 1, and measured data was analyzed. Unlike the first embodiment, the row board is composed of seven boards. The temperature was measured every hour for a period of about 500 days. The results are shown in FIGS.

  Each graph shown in FIGS. 10 to 12 is the same as FIGS. The window period for calculating the moving average value and the moving standard deviation is two weeks as in the first embodiment. As can be seen from FIGS. 10 and 11, the actually measured temperature data has a change of about 20 degrees or more (daily change and seasonal change) for any of the temperature sensors. The temperature sensor has a substantially constant value. As will be described later with reference to FIG. 12, similarly to the first embodiment, in the difference data, a long-term temperature change due to the season is canceled, but a relatively large daily temperature change remains. On the other hand, in the moving average value, the daily temperature change is canceled or reduced. Even in the moving standard deviation value, the daily temperature change is canceled or reduced.

  The vertical and horizontal axes of the graph shown in FIG. 12 are the same as those in FIG. The uppermost graph in FIG. 12 shows the average value of the measurement data of the seven temperature sensors as a representative value, as shown in the second graph from the top in the leftmost column of FIG. 10 (the measurement data of the temperature sensor in (2)). 9 shows difference data calculated from the measured data. As described above, in the difference data, a long-term temperature change due to the season is canceled, but a relatively large daily temperature change remains.

  The graphs at the second and third rows from the top in FIG. 12 are the same as the graphs at the center and right end of the second row from the top in FIG. The fourth and fifth graphs from the top in FIG. 12 show simulation results simulating the occurrence of abnormal heat generation. Specifically, 3 ° C. was added to each of the measurement data (A2) for 450 days or more counted from the start of the measurement (the period indicated by the arrow in FIG. 12) assuming the occurrence of abnormal heat generation. The results of calculating the moving average value and the moving standard deviation value using the data in the same manner as described above are the graphs in the fourth and fifth rows from the top in FIG. 12, respectively. As can be seen from a comparison between the second graph and the fourth graph showing the moving table average value, a clear change occurs in the moving table average value in the period in which 3 ° C. is added. Similarly, as can be seen from a comparison between the third graph and the fifth graph from the top showing the moving standard deviation value, in the period in which 3 ° C. is added, a clear change occurs in the moving standard deviation value. I have. As described above, even for a row board installed outdoors, the occurrence of temperature abnormality can be clearly detected by the moving average value and the moving standard deviation value calculated from the measurement data of the surface temperature of each board.

  In the above description, the measurement period is one hour, and the window period for calculating the moving average value and the moving standard deviation value is two weeks (14 days). However, the present invention is not limited to this. The measurement cycle may be shorter or longer than one hour. Further, the window period may be a period shorter or longer than 14 days.

  In the case of electrical facilities such as factories that are closed on Saturday and Sunday, the difference data often shows a regular trend every week. Therefore, in order to cancel the influence, a window period of 7 days ( It is preferable that the period be a natural number (a positive integer multiple) of one week). In addition, with regard to a row board arranged outdoors, the dispersion of the movement standard deviation value tends to increase due to the influence of the surrounding environment such as sunshine. In such a case, making the window period relatively long (for example, 21 days or 28 days) facilitates the determination.

  As described above, the present invention has been described by describing the embodiment. However, the above-described embodiment is an exemplification, and the present invention is not limited to the above-described embodiment. The scope of the present invention is shown by each claim of the claims, in consideration of the description of the detailed description of the invention, and all changes within the meaning and range equivalent to the language described therein are described. Including.

100 abnormality detection system 110 first temperature sensor 112 second temperature sensor 114 third temperature sensor 116 fourth temperature sensor 118 fifth temperature sensor 120 first switchboard 122 second switchboard 124 third switchboard 126 fourth switchboard 128 fifth switchboard 130 Data collection device 132 Analysis device 140, 160, 170 Control unit 142, 162, 172 Storage unit 144, 164, 174 Communication unit 146, 176 Timer 148, 166, 182 Bus 150 Sensor element 152 A / D conversion unit 178 Display unit 180 Operation unit 200 Sixth temperature sensor 202 Sixth switchboard 204 Seventh switchboard

Claims (5)

A first step of acquiring measurement data measured at the same timing from each of the plurality of sensors arranged in the electrical equipment;
A second step of determining a representative value from the plurality of acquired measurement data;
A third step of subtracting the representative value from each of the plurality of measurement data to calculate difference data;
Using the difference data obtained by repeatedly executing the first step, the second step, and the third step for a predetermined period, at least one of a moving average value and a moving standard deviation value is calculated for each sensor. A fourth step to
A step of comparing at least one of the calculated moving average value and the moving standard deviation value with a predetermined threshold value to determine whether an abnormality has occurred in the electrical equipment. Abnormality detection method. In the fourth step, the moving average value and the moving standard deviation value are calculated,
From the moving average value and the moving standard deviation value, calculating a deviation value of the difference data,
Comparing the absolute value of the divergence value with a predetermined threshold value to determine whether an abnormality has occurred in the electrical equipment,
L is a predetermined constant, the difference data of the i-th sensor among the plurality of sensors is Bi, and the difference within the predetermined period includes the difference data calculated before the difference data Bi as the latest difference data. The moving average calculated using the data is defined as AVBim, and the moving standard deviation calculated using the difference data within the predetermined period including the difference data calculated before the difference data Bi as the latest difference data. Assuming that the value is σBim, the deviation value Qi of the difference data Bi is
If (AVBim−L · σBim) ≦ Bi ≦ (AVBim + L · σBim), then Qi = 0,
If (AVBim−L · σBim)> Bi, then Qi = Bi− (AVBim−L · σBim), and
2. The abnormality detection method according to claim 1, wherein if Bi> (AVBim + L · σBim), Qi = Bi− (AVBim + L · σBim).   The abnormality detection method according to claim 1, wherein the predetermined period is a period that is a natural number multiple of 7 days. The sensor is a temperature sensor,
The abnormality detection method according to any one of claims 1 to 3, wherein the heights of the arrangement positions of the plurality of sensors on the electrical equipment are the same as each other. A plurality of sensors located in the electrical installation,
Data collection means for acquiring measurement data from each of the plurality of sensors at the same timing for a predetermined period,
Analysis means for analyzing the measurement data obtained by the data collection means,
The analysis means,
Determine a representative value from the plurality of measurement data measured at the same timing, calculate the difference data by subtracting the representative value from each of the plurality of measurement data measured at the same timing,
Using the difference data obtained in the predetermined period, for each of the sensors, calculate at least one of a moving average value and a moving standard deviation value,
An abnormality detection system that determines whether or not an abnormality has occurred in the electrical equipment by comparing at least one of the calculated moving average value and the calculated moving standard deviation value with a predetermined threshold value. .

Images