Monitoring and diagnostic system and monitoring and diagnostic method

Abstract

To provide a monitoring diagnostic system which enables use of the correlation anomaly detection method on a particular parameter even when the parameter has become perfectly constant in a target period.SOLUTION: A monitoring diagnostic system is provided, comprising an input unit for receiving input of time-series data on operation of a device, a determination unit configured to determine that the time-series data is abnormal when variation of the time-series data is less than a given amount during a certain period, and a correction unit configured to correct the time-series data when the data is determined to be abnormal by the determination unit.SELECTED DRAWING: Figure 8

Description

Technical Field

[0001] The present invention relates to a monitoring and diagnosis system and a monitoring and diagnosis method. Background Art

[0002] Conventionally, there has been known a method for determining whether equipment such as a chemical plant or a power plant is operating normally or an abnormality has occurred by using measured values such as temperature, pressure, vibration, and rotation speed measured in the equipment. Patent Document 1 describes a diagnosis method for diagnosing plant thermal efficiency by performing optimal state evaluation on measurement data for which high-precision measurement is difficult within the range of measurement accuracy and converging the data. Prior Art Documents Patent Documents

[0003] Patent Document 1 Japanese Unexamined Patent Publication No. 2002-122005 Summary of the Invention Problems to be Solved by the Invention

[0004] There exists a method of calculating a correlation coefficient between parameters and detecting whether equipment is operating normally or an abnormality has occurred based on the change in the correlation coefficient (hereinafter referred to as a correlation abnormality detection method). The correlation abnormality detection method has a problem that when a certain parameter becomes a completely constant value in a target section, the variance becomes 0, and division by zero occurs when calculating the correlation coefficient, so that the correlation coefficient cannot be calculated correctly.

[0005] An object of the present invention is to provide a monitoring and diagnosis system and a monitoring and diagnosis method that enable the correlation abnormality detection method to be used for a specific parameter even when the specific parameter becomes a completely constant value in a target section. Means for Solving the Problems

[0006] A monitoring and diagnostic system according to one aspect of the present invention includes an input unit for inputting time-series data relating to the operation of equipment, a determination unit for determining that the time-series data is abnormal if the variation in the time-series data is less than a predetermined amount for a certain period of time, and when the determination unit determines that the time-series data is abnormal By adding noise to the time-series data that is less than or equal to the measurement accuracy of the measuring instrument that measured the time-series data, The system includes a correction unit for correcting the aforementioned time-series data. A monitoring and diagnostic method according to one aspect of the present invention inputs time-series data relating to the operation of equipment, determines that the time-series data is abnormal if the variation in the time-series data is smaller than a predetermined amount for a certain period of time, and if it is determined that the data is abnormal, By adding noise to the time-series data that is less than or equal to the measurement accuracy of the measuring instrument that measured the time-series data, The aforementioned time-series data is corrected. [Effects of the Invention]

[0007] According to the present invention, even if a specific parameter becomes a completely constant value within the target interval, the correlation anomaly detection method can be used for that parameter. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 is a schematic diagram showing the overall configuration of the monitoring and diagnostic system according to the first embodiment. [Figure 2] Figure 2 is a schematic block diagram showing the hardware configuration of the monitoring and diagnostic device according to the first embodiment. [Figure 3] Figure 3 is a schematic block diagram showing the configuration of the monitoring and diagnostic device according to the first embodiment. [Figure 4] Figure 4 is a schematic diagram showing an example of time-series data input by the input unit. [Figure 5] Figure 5 is a schematic diagram showing an example of an instrument database stored in non-volatile memory. [Figure 6] Figure 6 illustrates the results of anomaly detection for time-series data and the values ​​before and after correction. [Figure 7] Figure 7 shows an example of a display screen. [Figure 8]Figure 8 is a flowchart of the monitoring and diagnostic process performed by the monitoring and diagnostic device. [Figure 9] Figure 9 illustrates the results of anomaly detection for time-series data and the values ​​before and after correction. [Modes for carrying out the invention]

[0009] (First Embodiment) A monitoring and diagnostic system according to an embodiment of the present invention will be described with reference to Figures 1 to 8.

[0010] Figure 1 is a schematic diagram showing the overall configuration of the monitoring and diagnostic system according to the first embodiment. The monitoring and diagnostic system 1 is a system that monitors and diagnoses whether the equipment 2 is operating normally using a monitoring and diagnostic device 3. The equipment 2 is, for example, a centrifugal compressor installed in a chemical plant or a power plant. Multiple measuring instruments 4 are installed near the equipment 2 to measure various parameters of the equipment 2 itself or the surrounding environment in which the equipment 2 is installed. The parameters measured by the measuring instruments 4 include, for example, the rotational speed of the prime mover in the equipment 2, the pressure inside the cylinder of the prime mover, the internal temperature of the prime mover, and the ambient temperature around the equipment 2.

[0011] The measuring instrument 4 and the monitoring and diagnostic device 3 are connected to the network 6 by wire or wireless connection, enabling data communication from the measuring instrument 4 to the monitoring and diagnostic device 3. The measuring instrument 4 measures various parameters as described above and outputs the measured values ​​as time-series data. The time-series data output from multiple measuring instruments 4 are input to the monitoring and diagnostic device 3 via the network 6. The monitoring and diagnostic device 3 uses these multiple time-series data to diagnose the operating status of the equipment 2 and monitors whether the equipment 2 is operating normally. The monitoring and diagnostic device 3 outputs the result of the abnormality determination of the equipment 2 to the display device 5. The display device 5 is a display device equipped with, for example, a liquid crystal display unit. The display device 5 is attached to the monitoring and diagnostic device 3 or installed outside the monitoring and diagnostic device 3. Users of the monitoring and diagnostic device 3 can know the operating status of the equipment 2 by visually checking the display on the display device 5. In other words, the monitoring and diagnostic device 3 notifies users of the monitoring and diagnostic device 3 of the diagnostic result of the equipment 2 by outputting the result of the abnormality determination to the display device 5.

[0012] Figure 2 is a schematic block diagram showing the hardware configuration of the monitoring and diagnostic device 3 according to the first embodiment. The monitoring and diagnostic device 3 consists of a computer equipped with a processing unit 10 such as a CPU (Central Processing Unit), MPU (Micro Processing Unit), and DSP (Digital Signal Processor), non-volatile memory 11 such as ROM (Read Only Memory), flash memory, and hard disk drive, volatile memory 12 known as RAM (Random Access Memory), an input / output interface 13, and other peripheral circuits. These hardware components work together to operate the software and realize multiple functions. The monitoring and diagnostic device 3 may be composed of one computer or multiple computers. Furthermore, the processing unit 10 can be an ASIC (application specific integrated circuit), FPGA (Field Programmable Gate Array), etc.

[0013] The non-volatile memory 11 stores a program capable of executing various calculations. That is, the non-volatile memory 11 is a readable storage medium (storage device) that stores a program for realizing the functions of the present embodiment. The volatile memory 12 is a storage medium (storage device) that temporarily stores calculation results obtained by the processing device 10 and signals input from the input / output interface 13. The processing device 10 is a device that loads the program stored in the non-volatile memory 11 into the volatile memory 12 and executes calculations, and performs predetermined calculation processing on data acquired from the input / output interface 13, the non-volatile memory 11 and the volatile memory 12 in accordance with the program.

[0014] In addition to the aforementioned program, the non-volatile memory 11 stores a measuring instrument database 30 and comparison data 39. Details of the measuring instrument database 30 and the comparison data 39 will be described later.

[0015] The input / output interface 13 is connected to the network 6 and the display device 5 respectively. An input unit of the input / output interface 13 converts signals input from a plurality of measuring instruments 4 via the network 6 into data that can be calculated by the processing device 10. An output unit of the input / output interface 13 generates an output signal according to the calculation result obtained by the processing device 10, and outputs the signal to the display device 5.

[0016] FIG. 3 is a block diagram schematically showing the configuration of the monitoring and diagnostic device 3 according to the first embodiment. The monitoring and diagnostic device 3 includes an input unit 20, a determination unit 21, a correction unit 22, a diagnostic unit 23, and an output unit 24. Each of these functional units is functionally implemented by a program stored in the non-volatile memory 11.

[0017] The input unit 20 receives time-series data relating to the operation of the device 2. Specifically, the input unit 20 receives a plurality of pieces of time-series data respectively measured by a plurality of measuring instruments 4. The determination unit 21 determines whether the input time-series data is normal or abnormal. The correction unit 22 corrects specific time-series data when the determination unit 21 determines that the specific time-series data is abnormal. The diagnosis unit 23 determines whether the operation of the device 2 is normal or abnormal using the time-series data corrected by the correction unit 22 and the time-series data not corrected by the correction unit 22 (that is, the time-series data as input by the input unit 20). The output unit 24 outputs the result of the determination by the determination unit 21 to the display device 5.

[0018] Figure 4 is a schematic diagram showing an example of time-series data input by the input unit 20. In Figure 4, four pieces of time-series data respectively corresponding to four parameters, namely rotation speed, pressure A, temperature A, and temperature B, are collectively illustrated as one example. Each piece of time-series data is associated with a measurement date and time 41.

[0019] Figure 5 is a schematic diagram showing an example of the measuring instrument database 30 stored in the non-volatile memory 11. For each of the measuring instruments 4, the measuring instrument database 30 stores items of a sensor ID 31, a parameter 32, a unit 33, a range lower limit 34, a range upper limit 35, and a measurement accuracy 36. The sensor ID 31 is a unique identifier for distinguishing each individual measuring instrument 4. The parameter 32 is a parameter measured by the measuring instrument 4. The unit 33 is the unit of the measurement value measured by the measuring instrument 4.

[0020] The lower range limit 34 is the lower limit of the value that the measuring instrument 4 can measure. For example, if the lower range limit 34 of a measuring instrument 4 that measures temperature is "0.0", it means that the measuring instrument 4 can only measure temperatures down to 0°C, and when the temperature is below 0°C (for example, a value below freezing such as -2°C), the measuring instrument 4 will output a measured value of "0.0". The upper range limit 35 is the upper limit of the value that the measuring instrument 4 can measure. For example, if the upper range limit 35 of a measuring instrument 4 that measures temperature is "100.0", it means that the measuring instrument 4 can only measure temperatures up to 100°C, and when the temperature is 100°C or higher (for example, a value such as 120°C), the measuring instrument 4 will output a measured value of "100.0". The measurement accuracy 36 indicates the resolution of the measuring instrument 4. For example, a measurement accuracy of 36 of "0.1" means that the measurement value of the measuring instrument 4 can be trusted in units of 0.1, but not in units smaller than 0.1. In other words, in this case, the value "10.4999" output by the measuring instrument 4 can be trusted up to "10.4", but the smaller part of the value, "0.0999", cannot be trusted.

[0021] Figure 6 illustrates the results of anomaly detection for time-series data and the values ​​before and after correction. In Figure 6, for a parameter called "Temperature A," which is the temperature of a specific part of the device 2, the measurement date and time 51, the measured value 52, the anomaly detection result 53, and the correction value 54 are shown. For example, the first row of Figure 6 shows the measurement date and time 51, "1 / 1 23:58," the corresponding measured value 52, "22.56℃," the anomaly detection result 53, "Normal," and the correction value 54, "22.56℃." The measurement date and time 51 is the date and time when the measuring instrument 4 performed the measurement. The measured value 52 is the measured value obtained as a result of the measurement by the measuring instrument 4. Since the parameter exemplified in Figure 6 is related to temperature, the unit of the measured value 52 is "℃." The anomaly detection result 53 is the result of the anomaly detection processing by the detection unit 21. The anomaly detection result 53 takes one of two values: "Normal" or "Abnormal." The corrected value 54 is the measured value 52 itself if the abnormality judgment result 53 is "normal," and is the value corrected by the correction unit 22 if the abnormality judgment result 53 is "abnormal." In the following explanation, this corrected value 54, which has been corrected according to the abnormality judgment result 53, will be referred to as the corrected time series data.

[0022] Referring to Figure 6, the abnormality detection process of time-series data by the determination unit 21 will be explained. The determination unit 21 reads the upper and lower limits of the measurement range of the measuring instrument 4 from the measuring instrument database 30 (see Figure 5). Let's assume that the upper limit of the measurement range of the measuring instrument 4 that measures "temperature A" is 100.0℃ and the lower limit is 20.0℃. In Figure 6, during the period from "1 / 2 0:00" to "1 / 2 23:59", the actual value of "temperature A" is below the lower limit of the measurement range, i.e., 20.0℃. However, since the lower limit of the measurement range is 20.0℃, the measuring instrument 4 outputs 20.0℃ as the measured value of "temperature A". The determination unit 21 determines that if five or more consecutive measured values ​​are exactly the same as the lower or upper limit of the measurement range (i.e., if the variability of the time series data is less than a predetermined amount for a certain period), the measured values ​​for that period have deviated from the measurement range, and the time series data for that period is determined to be abnormal data. In the example in Figure 6, the determination unit 21 determines that the time series data for the period from "1 / 2 0:00" to "1 / 2 23:59" is abnormal data.

[0023] Note that the measured value that the determination unit 21 determines to be outside the measurement range does not have to be exactly the same as the lower or upper limit of the measurement range. For example, if the measuring instrument 4 is not operating correctly, it may be configured to output a predetermined reserved value (a value that is guaranteed not to be output during normal operation). In such a case, if the reserved value is output continuously, the determination unit 21 determines that the time-series data for that period is abnormal data. Examples of when the measuring instrument 4 is not operating correctly include cases where the voltage supplied to the measuring instrument 4 is insufficient, or where the wiring connecting the measuring instrument 4 to the peripheral circuit is broken or short-circuited.

[0024] The correction process performed by the correction unit 22 will now be explained. The correction unit 22 corrects the data from among the multiple time series data input by the input unit 20 for a period in which the measured values ​​are exactly the same as the lower limit or upper limit of the measurement range for five or more consecutive points. In other words, the correction unit 22 corrects the time series data from among the multiple time series data input by the input unit 20 for which the variability is less than a predetermined amount for a certain period of time. For example, the correction unit 22 reads the measurement accuracy of the measuring instrument 4 from the measuring instrument database 30 and adds random noise of a normal distribution with the measurement accuracy as the standard deviation to the measured values ​​for the above period.

[0025] If the measured values ​​remain exactly the same for a certain period, the variance for that period becomes zero, which can lead to a division-by-zero error when calculating the correlation coefficient. By adding random noise in this way, the repetition of exactly the same values ​​is prevented, so the variance for that period is no longer zero, and the correlation coefficient can be calculated without the risk of a division-by-zero error. If the noise is at the level of the measurement accuracy, the impact of the correction on the correlation coefficient is sufficiently small and does not pose a problem. Note that instead of generating random noise using the measurement accuracy of instrument 4 itself as the standard deviation, random noise below the measurement accuracy may be generated.

[0026] It should be noted that even if corrections are made for abnormal constant values, correct measurement values ​​cannot be obtained. However, by first enabling calculation using the correlation anomaly detection method, it is possible to make a rough judgment as to whether the monitoring and diagnosis of device 2 is effective. In the conventional method, it was necessary to completely eliminate the abnormal constant values, but the method of this embodiment can reduce that effort.

[0027] The diagnostic unit 23 compares the comparison data 39 stored in the non-volatile memory 11 with the corrected time-series data to determine whether the operation of the device 2 is normal or abnormal. The comparison data 39 is time-series data measured by the measuring instrument 4 at a point in the past when the device 2 was operating normally. The diagnostic unit 23 calculates the correlation coefficient between the comparison data 39 and the corrected time-series data, and determines that the operation of the device 2 is abnormal if the correlation coefficient of the latter has changed by a certain amount or more relative to the correlation coefficient of the former (a so-called correlation anomaly detection method).

[0028] Examples of abnormalities in equipment 2 detected by a correlation anomaly detection method using the correlation of measured data are given below. For example, there is a correlation between the rotational speed and flow rate of a pump. Under normal circumstances, the correlation coefficient for the rotational speed of a pump is 0.9, but if the flow rate is low relative to the rotational speed for a certain period, the correlation coefficient will drop to a smaller value such as 0.8. The diagnostic unit 23 can detect an abnormality in the pump flow rate parameter from such a change in the correlation coefficient and detect a pump abnormality. Similarly, for example, there is a correlation between the rotational speed and vibration of a turbine, and the temperature of multiple bearings of the turbine, and abnormalities can be detected using the correlation anomaly detection method.

[0029] Figure 7 shows an example of the screen display of the display device 5. The output unit 24 outputs data including the result of the determination by the determination unit 21 to the display device 5. Based on the data input from the output unit 24, the display device 5 displays an image on the display screen 60. The display screen 60 exemplified in Figure 7 displays the anomaly degree 61 of the time series data and the period 62 of the time series data that the determination unit 21 determined to be an anomaly.

[0030] Figure 8 is a flowchart of the monitoring and diagnostic process performed by the monitoring and diagnostic device 3. In step S100, the input unit 20 receives multiple time-series data measured by multiple measuring instruments 4. In step S110, the determination unit 21 determines whether the multiple time-series data input in step S100 contains five or more consecutive measured values ​​that are exactly the same as the lower limit or upper limit of the measurement range of the measuring instrument 4, that is, whether the time-series data is abnormal. If at least one of the time-series data is determined to be abnormal, the process proceeds to step S120. In step S120, the correction unit 22 corrects the measured values ​​corresponding to the period determined to be abnormal among the time-series data that was determined to be abnormal. After that, the process proceeds to step S130. On the other hand, if all the time-series data is determined to be normal in step S110, the process proceeds to step S130.

[0031] In step S130, the diagnostic unit 23 performs a correlation anomaly detection method to determine whether the operation of the device 2 is abnormal. Specifically, the diagnostic unit 23 calculates the correlation coefficient between the comparison data 39 and the corrected time series data, and determines whether the correlation coefficient of the latter has changed by a certain amount or more compared to the correlation coefficient of the former (correlation coefficient under normal conditions). In step S140, the output unit 24 outputs the result of the determination performed by the determination unit 21 in step S110 to the display device 5.

[0032] As described above, in the monitoring and diagnostic method according to this embodiment, the monitoring and diagnostic device 3 inputs time-series data relating to the operation of the device 2 (S100), determines that the time-series data is abnormal if the variation in the time-series data is smaller than a predetermined amount for a certain period of time (Yes in S110), and corrects the time-series data if it is determined to be abnormal (S120).

[0033] According to the first embodiment described above, the following effects are achieved.

[0034] (1) The determination unit 21 determines that the time series data is abnormal if the variation in the time series data relating to the operation of the device 2 input by the input unit 20 is zero (i.e., smaller than a predetermined amount) for a period of 5 points (a certain period). The correction unit 22 then corrects the time series data if the determination unit 21 has determined that it is abnormal. In this way, even if a particular parameter becomes a completely constant value in the target interval, it becomes possible to use the correlation anomaly detection method for that parameter.

[0035] (2) The input unit 20 receives multiple time series data, and the correction unit 22 corrects the time series data from which the variation is zero (less than a predetermined amount) for a period of 5 points (a certain period). In this way, time series data in which no abnormalities are observed are used as is, and no extraneous noise is added, so the accuracy does not decrease.

[0036] (3) The correction unit 22 corrects the data in the time series data that corresponds to the portion where the measured value is a constant value for a period of 5 points (a certain period). In this way, only the abnormal constant values ​​are corrected, and the rest of the time series data does not need to be corrected, so no extraneous noise is added and the accuracy does not decrease.

[0037] (4) The correction unit 22 corrects the time series data by adding random noise with a measurement accuracy of 36 or less from the measuring instrument 4 that measured the time series data. In this way, the variance of the time series data will never be zero, and no division by zero error will occur even if the correlation anomaly detection method is applied.

[0038] (5) The output unit 24 outputs the result of the determination by the determination unit 21 to the display device 5 (external device). In this way, the determination result is notified to the user, so the user can refer to the determination result to find signs of abnormality in the device 2 or measuring instrument 4.

[0039] (Second embodiment) Referring to Figure 9, a monitoring and diagnostic system according to the second embodiment of the present invention will be described. Note that the same or equivalent components as those described in the first embodiment will be denoted by the same reference numerals, and the differences will be primarily explained.

[0040] Figure 9 is similar to Figure 6 and illustrates the results of anomaly detection for time-series data and the values ​​before and after correction. The measured value 52 shown in Figure 9 relates to "temperature B". In the second embodiment, the determination unit 21 determines anomalies in the time-series data using the measurement accuracy rather than the upper and lower limits of the measurement range of the measuring instrument 4.

[0041] In the example shown in Figure 9, the measured value 52, which represents temperature, is a constant value of approximately 20°C at measurement time 51, which is after "1 / 2 0:00". Here, the measurement range is not deviated. However, at measurement time 51, "1 / 2 0:03", a measured value 52 of "19.9999" is obtained, so it cannot be said that the measured value 52 is a completely constant value after "1 / 2 0:00".

[0042] Such time-series data can be seen when, for some reason, correct measurements are not being taken, such as a broken wire in the measuring instrument 4, resulting in a constant value being output, and including variations of a floating-point error. In such time-series data, the measured value 52 is almost constant, but there are slight changes, making it difficult to determine whether it is an abnormal constant value by checking whether the measured value 52 is perfectly constant, as in the first embodiment. However, as shown in Figure 5, the measurement accuracy 36 of "temperature B" is "0.1", so it is unlikely that the actual value can be measured with high accuracy to four decimal places, such as "19.9999".

[0043] In the second embodiment, the determination unit 21 calculates the variability of the time series data using the standard deviation of the five most recent measurement points. If this is smaller than the measurement accuracy 36, it determines that the time series data contains an abnormal constant value (i.e., the time series data is abnormal). In the example shown in Figure 9, the determination unit 21 determines that the measurement value from "1 / 2 0:04" is an abnormal constant value.

[0044] In the second embodiment, the measurement accuracy 36 included in the instrument database 30 is used to mean that there is at least this much measurement error. For example, if the measurement accuracy 36 is the allowable accuracy of the instrument 4, that is, a value that means that if the error exceeds this, the instrument 4 needs to be calibrated, then it is desirable to consider 1 / 10 of the allowable accuracy as the measurement accuracy of the instrument 4. Also, if there is no information on the measurement accuracy of the instrument 4 and the measurement accuracy 36 cannot be stored in the instrument database 30, the determination unit 21 can perform abnormality determination processing by considering, for example, (upper measurement range - lower measurement range) × 0.1% or (maximum measurement data - minimum measurement data) × 0.1% as the measurement accuracy. In this way, the correction unit 22 corrects the time series data based on the difference between the upper and lower measurement limits of the instrument 4 that measured the time series data, or the difference between the maximum and minimum values ​​of the data included in the time series data, so that the time series data can be appropriately corrected even if the measurement accuracy is unknown.

[0045] According to the second embodiment described above, the following effects are achieved.

[0046] The determination unit 21 determines that the time-series data is abnormal if the standard deviation (variability) of the five measured values ​​of the time-series data relating to the operation of the device 2 input by the input unit 20 is smaller than the measurement accuracy 36 (a predetermined amount) for a period of five points (a certain period). Then, the correction unit 22 corrects the time-series data if the determination unit 21 has determined that it is abnormal. In this way, even if a particular parameter has almost the same constant value in the target interval, it becomes possible to use the correlation anomaly detection method for that parameter.

[0047] The following modifications are also within the scope of the present invention, and it is possible to combine the configurations shown in the modifications with the configurations described in the embodiments described above, or to combine the configurations described in the different embodiments described above, or to combine the configurations described in the following different modifications.

[0048] (Variation 1) The correction unit 22 may perform corrections not only for the period determined to be abnormal by the determination unit 21, but also for other periods. By doing so, the presence or absence of random noise can be made common between the period determined to be abnormal and the other periods. In other words, the preconditions for the correction value can be made consistent for the entire period of the time series data. In certain types of processing, such unification of preconditions can be advantageous. Furthermore, not only the parameter (time series data) that includes the period determined to be abnormal by the determination unit 21 may be subject to correction, but all other parameters (all other time series data) may also be subject to correction.

[0049] (Modification 2) In the first embodiment, consecutive identical measurement values ​​are used to determine abnormalities, while in the second embodiment, the smallness of the variation due to the standard deviation is used. However, both may be examined to determine abnormalities. In this case, if an abnormality is detected in either one, the correction unit 22 will perform a correction process.

[0050] (Variation 3) In the first embodiment, consecutive identical measurement values ​​are used to determine abnormality, and in the second embodiment, the smallness of the variation due to the standard deviation is used. However, it is also possible to check whether the variation in time-series data is smaller than a predetermined amount for a certain period of time using methods other than these. For example, it is possible to determine that an abnormal constant value exists if the difference between the most recent measurement value and the current measurement value is less than or equal to the measurement accuracy for 10 consecutive points.

[0051] (Modification 4) The method of correcting time-series data by the correction unit 22 is not limited to that described in the first embodiment. For example, the correction unit 22 may add random noise with a normal distribution using the standard deviation of the comparison data 39. Alternatively, the correction unit 22 may add sine wave or step-like noise with amplitude equal to the standard deviation of the previous 10 measured values. Furthermore, the correction unit 22 may add noise equal to 0.1% of the measurement range (or the difference between the maximum and minimum measured values ​​included in the time-series data).

[0052] Although embodiments of the present invention have been described above, these embodiments only represent a part of the application examples of the present invention, and are not intended to limit the technical scope of the present invention to the specific configurations of the above embodiments. [Explanation of Symbols]

[0053] 1...Monitoring and diagnostic system, 2...Equipment, 3...Monitoring and diagnostic device, 4...Measuring instrument, 5...Display device, 6...Network, 10...Processing device, 11...Non-volatile memory, 12...Volatile memory, 13...Input / output interface, 20...Input unit, 21...Determination unit, 22...Correction unit, 23...Diagnostic unit, 24...Output unit, 30...Measuring instrument database, 39...Comparison data

Claims

1. An input section for inputting time-series data related to the operation of the device, A determination unit that determines that the time series data is abnormal if the variation in the time series data is less than a predetermined amount for a certain period of time, If the determination unit determines that an abnormality has occurred, the correction unit corrects the time-series data by adding noise to the time-series data that is less than or equal to the measurement accuracy of the measuring instrument that measured the time-series data. A monitoring and diagnostic system equipped with the following features.

2. In the monitoring and diagnostic system according to claim 1, The input unit receives a plurality of the aforementioned time-series data, The correction unit is a monitoring and diagnostic system that corrects time series data from among the multiple time series data for which the variation is smaller than a predetermined amount over a certain period of time.

3. In the monitoring and diagnostic system according to claim 1, The correction unit is a monitoring and diagnostic system that corrects the data included in the time-series data that corresponds to the specified period.

4. In the monitoring and diagnostic system according to claim 1, The correction unit is a monitoring and diagnostic system that corrects the time-series data by adding random noise as noise to the time-series data.

5. In the monitoring and diagnostic system according to claim 1, The correction unit corrects the time series data by adding noise equal to 0.1% of the difference between the maximum and minimum values ​​of the data included in the time series data to the time series data, thereby correcting the time series data.

6. In the monitoring and diagnostic system according to claim 1, A monitoring and diagnostic system further comprising an output unit that outputs the result of the determination by the determination unit to an external device.

7. Input time-series data related to the operation of the equipment. If the variability of the aforementioned time series data is less than a predetermined amount for a certain period of time, the aforementioned time series data is determined to be abnormal. If the aforementioned abnormality is determined, the time-series data is corrected by adding noise to the time-series data that is below the measurement accuracy of the measuring instrument that measured the time-series data. Monitoring and diagnostic methods.

Images