Sign detection device, sign detection system, and sign detection method

Abstract

To improve accuracy of sign detection of cooling performance deterioration.SOLUTION: A sign detection device includes an acquisition unit that acquires time-series data indicating an operation status of a cooling target device including a heat source and a cooling unit that cools the heat source with a cooling medium and including a temperature of the cooling medium, and a sign detection unit that detects a sign of deterioration in cooling performance of the cooling unit using the time-series data acquired by the acquisition unit and a sign detection model. The sign detection model is constructed using a time-series feature amount of specific time-series data in which a temperature of the cooling medium is equal to or higher than a threshold value among the time-series data and a label indicating normality or abnormality of cooling performance of the cooling unit.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a sign detection device, a sign detection system, and a sign detection method for detecting signs. [Background technology]

[0002] The oil cooling device is installed within the equipment being cooled, and as long as the equipment is operating within its rated range, it cools the oil within the equipment to an appropriate temperature range. This prevents the equipment from reaching an abnormal state. However, if the cooling performance of the oil cooling device deteriorates, the oil temperature within the equipment will rise, causing the equipment to reach an abnormal state, making safe and stable operation of the equipment difficult. Normally, if the oil temperature rises and reaches an abnormal state, the oil cooling device's check mechanism will activate and the equipment will stop operating.

[0003] Furthermore, Patent Document 1 below discloses a failure sign diagnosis system that is configured with a diagnostic execution unit, a deployment unit, a device to be diagnosed, a diagnostic server, and a network. In this failure sign diagnosis system, the diagnostic execution unit has processing modules for sensor input processing, pre-processing, diagnostic processing, and post-processing, and a common interface that connects the processing modules, and the deployment unit deploys and executes the processing modules in the device to be diagnosed or the diagnostic server. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-12157 Summary of the Invention [Problem to be solved by the invention]

[0005] Factors that cause deterioration in the cooling performance of an oil cooling system include clogged oil coolers, clogged oil filters, deterioration in oil quality, and deterioration of oil piping. Deterioration in the cooling performance of an oil cooling system appears as a difference in oil temperature for the same load on the equipment being cooled. However, when the load is high, a significant difference in oil temperature appears, but when the load is low, the difference in oil temperature does not appear significantly, making it difficult to detect signs of deterioration in cooling performance.

[0006] An object of the present invention is to improve the accuracy of detecting signs of cooling performance deterioration. [Means for solving the problem]

[0007] An early warning detection device, which is one aspect of the invention disclosed in the present application, comprises an acquisition unit that acquires time-series data indicating the operating status of a cooled device having a heat source and a cooling unit that cools the heat source with a cooling medium and including the temperature of the cooling medium, and an early warning detection unit that performs early detection of deterioration in the cooling performance of the cooling unit using the time-series data acquired by the acquisition unit and an early warning detection model, wherein the early warning detection model is constructed using time-series features of specific time-series data from the time-series data where the temperature of the cooling medium is above a threshold value and a label indicating whether the cooling performance of the cooling unit is normal or abnormal. Another aspect of the invention disclosed in this application is a symptom detection device that includes an acquisition unit that acquires time-series data that indicates the operating status of a cooling target device having a heat source and a cooling unit that cools the heat source with a cooling medium and that includes the temperature of the cooling medium; a pre-processing unit that calculates, as time-series features, specific time-series data from the time-series data acquired by the acquisition unit where the temperature of the cooling medium is equal to or higher than a threshold; and a symptom detection unit that performs symptom detection of deterioration in the cooling performance of the cooling unit using the time-series features calculated by the pre-processing unit and a symptom detection model. [Effects of the Invention]

[0008] According to the exemplary embodiment of the present invention, it is possible to improve the accuracy of detecting signs of cooling performance degradation. Problems, configurations, and effects other than those described above will become clear from the following description of the examples. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a block diagram showing an example of the system configuration of a warning sign detection system. [Figure 2] FIG. 2 is a block diagram showing a detailed system configuration example 1 of the symptom detection system. [Figure 3] FIG. 3 is a block diagram showing a detailed system configuration example 2 of the symptom detection system. [Figure 4] FIG. 4 is a block diagram showing an example of the configuration of the device to be cooled. [Figure 5] FIG. 5 is a block diagram showing an example of the hardware configuration of a computer. [Figure 6] FIG. 6 is an explanatory diagram illustrating an example of the sensor data table. [Figure 7] FIG. 7 is an explanatory diagram illustrating an example of a training data set table. [Figure 8] FIG. 8 is a graph showing time series data of the discharge temperature and the ambient temperature. [Figure 9] FIG. 9 is an explanatory diagram illustrating a first example of generation of a sign detection model. [Figure 10] FIG. 10 is an explanatory diagram showing a second example of generation of a sign detection model. DETAILED DESCRIPTION OF THE INVENTION

[0010] <Example of system configuration for early detection system> 1 is a block diagram showing an example of the system configuration of a symptom detection system. The symptom detection system 100 includes a cooling target device 101, a sampling processing unit 102, a data preprocessing unit 103, a construction unit 104, and a symptom detection unit 105.

[0011] The device to be cooled 101 has a heat source 111, a sensor 112, and an oil cooling unit 113. The heat source 111 is a source of heat generation within the device to be cooled 101, and if the device to be cooled 101 is, for example, an air compressor, it is a motor. The sensor 112 detects various operating conditions within the device to be cooled 101. The sensor 112 is, for example, a temperature sensor, an ammeter, or a pressure sensor. The oil cooling unit 113 is a mechanism that cools the oil circulating within the device to be cooled 101. In this embodiment, oil is used as an example of a cooling medium; however, the type of cooling medium to be used depends on the type of device to be cooled 101, and therefore a cooling medium other than oil, such as water or chlorofluorocarbon, may also be used.

[0012] The sampling processing unit 102 converts analog data from the sensor 112 into digital data and outputs it as sensor data 114 .

[0013] The data preprocessing unit 103 removes outliers from the sensor data 114 and interpolates the sensor data 114 at missing times, and outputs the resulting data as feature quantities 115 .

[0014] The construction unit 104 uses the features 115 and the positive / negative labels 116 as a training dataset and constructs the sign detection model 117. Specifically, for example, the construction unit 104 uses the training dataset to generate the sign detection model 117 by, for example, a decision tree, a random forest, or deep learning.

[0015] The sign detection unit inputs the feature amount 115 into a sign detection model 117 to output a diagnosis result 118 that indicates a sign of deterioration in the cooling performance of the oil cooling unit 113 .

[0016] 2 is a block diagram showing a detailed system configuration example 1 of the sign detection system 100. The sign detection system 100 has a user site 201, an operation site 202, and a cloud site 203. The user site 201 and the cloud site 203, and the operation site 202 and the cloud site 203 are communicatively connected via a network such as the Internet, a LAN (Local Area Network), or a WAN (Wide Area Network).

[0017] The user site 201 includes the cooled device 101 and a first communication control unit 210. In Fig. 2, the sampling processing unit 102 is included in the cooled device 101, but it may be outside the cooled device 101 as long as it is within the user site 201.

[0018] The production site 202 includes a data preprocessing unit 103 , a construction unit 104 , and a second communication control unit 220 .

[0019] The cloud site 203 includes a data preprocessing unit 103, a sign detection unit 105, and a third communication control unit 230.

[0020] First, a description will be given of the process of building the sign detection model 117 in system configuration example 1. At the user site 201, the cooling target device 101 outputs analog data detected by the sensor 112 to the sampling processing unit 102, and the sampling processing unit 102 outputs sensor data 114 to the first communication control unit 210. The user site 201 transmits the sensor data 114 to the third communication control unit 230 of the cloud site 203 via the first communication control unit 210.

[0021] The cloud site 203 transfers the sensor data 114 from the user site to the second communication control unit 220 of the production site 202 by the third communication control unit 230 .

[0022] At the operation site 202, the data preprocessing unit 103 acquires the sensor data 114 received by the second communication control unit 220 and outputs the features 115 to the construction unit 104. The construction unit 104 uses the training dataset (the features 115 and the positive / negative labels 116) to construct a sign detection model 117 and outputs it to the second communication control unit 220. The second communication control unit 220 transmits the sign detection model 117 to the third communication control unit 230 at the cloud site 203.

[0023] At the cloud site 203 , the third communication control unit 230 outputs the sign detection model 117 from the production site 202 to the sign detection unit 105 .

[0024] Next, a description will be given of the symptom detection process using the symptom detection model 117 in system configuration example 1. At the user site 201, the cooling target device 101 outputs analog data detected by the sensor 112 to the sampling processing unit 102, and the sampling processing unit 102 outputs sensor data 114 to the first communication control unit 210. The user site 201 transmits the sensor data 114 to the third communication control unit 230 of the cloud site 203 via the first communication control unit 210.

[0025] At the cloud site 203, the third communication control unit 230 outputs the sensor data 114 from the user site 201 to the data pre-processing unit 103. The data pre-processing unit 103 removes outliers from the sensor data 114 and interpolates the sensor data 114 at missing times, thereby outputting feature amounts 115 to the sign detection unit 105. The sign detection unit 105 inputs the feature amounts 115 to a sign detection model 117 and outputs a diagnosis result 118 that indicates a sign of deterioration in the cooling performance of the oil cooling unit 113.

[0026] 3 is a block diagram showing a detailed system configuration example 2 of the sign detection system 100. The following description will focus on the differences from system configuration example 1 in FIG.

[0027] The user site 201 includes a cooled device 101 and a first communication control unit 210. In Fig. 3, the sampling processing unit 102 and the data preprocessing unit 103 are included in the cooled device 101, but they may be outside the cooled device 101 as long as they are within the user site 201.

[0028] The production site 202 includes the construction unit 104 and a second communication control unit 220 .

[0029] The cloud site 203 includes a sign detection unit 105 and a third communication control unit 230.

[0030] In system configuration example 2, the data preprocessing unit 103 exists only at the user site 201. That is, by generating the feature 115 at the user site 201, it becomes possible to construct the sign detection model 117 and perform sign detection using sensor data 114 with a sampling period shorter than the sampling period of the sensor data 114 in system configuration example 1.

[0031] First, the process of building the sign detection model 117 in system configuration example 2 will be described. At the user site 201, the cooling target device 101 outputs analog data detected by the sensor 112 to the sampling processing unit 102, and the sampling processing unit 102 outputs sensor data 114 to the data pre-processing unit 103. The data pre-processing unit 103 removes outliers from the sensor data 114 and interpolates the sensor data 114 at missing times, thereby outputting feature amounts 115 to the first communication control unit 210. The user site 201 transmits the feature amounts 115 to the third communication control unit 230 of the cloud site 203 via the first communication control unit 210.

[0032] The cloud site 203 transfers the feature amount 115 from the user site to the second communication control unit 220 of the production site 202 by the third communication control unit 230 .

[0033] At the operation site 202, the second communication control unit 220 outputs the features 115 from the user site 201 to the construction unit 104. The construction unit 104 constructs a sign detection model 117 using the training dataset (the features 115 and the positive / negative labels 116), and outputs the model to the second communication control unit 220. The second communication control unit 220 transmits the sign detection model 117 to the third communication control unit 230 at the cloud site 203.

[0034] At the cloud site 203 , the third communication control unit 230 outputs the sign detection model 117 from the production site 202 to the sign detection unit 105 .

[0035] Next, the symptom detection process using the symptom detection model 117 in system configuration example 2 will be described. At the user site 201, the cooling target device 101 outputs analog data detected by the sensor 112 to the sampling processing unit 102, and the sampling processing unit 102 outputs sensor data 114 to the data pre-processing unit 103. The data pre-processing unit 103 removes outliers from the sensor data 114 and interpolates the sensor data 114 at missing times, thereby outputting feature amounts 115 to the first communication control unit 210. The user site 201 transmits the feature amounts 115 to the third communication control unit 230 of the cloud site 203 via the first communication control unit 210.

[0036] At the cloud site 203, the third communication control unit 230 outputs the feature amount 115 from the user site 201 to the sign detection unit 105. The sign detection unit 105 inputs the feature amount 115 to a sign detection model 117 and outputs a diagnosis result 118 that indicates a sign of deterioration in the cooling performance of the oil cooling unit 113.

[0037] <Configuration example of the device to be cooled 101> FIG. 4 is a block diagram showing an example of the configuration of the apparatus to be cooled 101. In FIG. 4, an air compressor is used as an example of the apparatus to be cooled 101. The apparatus to be cooled 101 includes an inverter 400, a heat source 111 such as a motor, a compressor 401, an oil pan 402, a check valve 403, an oil cooler 404, an oil pump 405, an oil filter 406, an aftercooler 407, and an air cooler 408. The inverter 400 controls the rotation of the motor, which is the heat source 111. As the frequency of the AC voltage converted by the inverter 400 increases, the load on the motor increases, causing the motor to rotate at a higher speed, resulting in the compressor 401 generating more compressed air. The apparatus to be cooled 101 also includes a first intake port 410, a second intake port 461, and an exhaust port 480.

[0038] The apparatus to be cooled 101 also has sensors 112, including an ammeter 411, a pressure gauge 451, a discharge thermometer 452, and an ambient thermometer 462. The ammeter 411 detects the current value of the heat source 111. The pressure gauge 451 detects the discharge pressure of the oil. The discharge thermometer 452 detects the discharge temperature of the oil. The ambient thermometer 462 detects the ambient temperature of the apparatus to be cooled 101 using the air from the second intake port 461. The sensor 112 also detects the voltage frequency of the inverter 400. The analog data output from each sensor 112 is sampled at the same time by the sampling processing unit 102.

[0039] The oil circulation path of the oil cooling unit 113 is the heat source 111 ⇒ oil pan 402 ⇒ check valve 403 ⇒ oil cooler 404 ⇒ oil pump 405 ⇒ oil filter 406 ⇒ heat source 111 ⇒ . . .

[0040] In addition, the air flows through the first intake port 410 ⇒ heat source 111 ⇒ compression section 401 ⇒ aftercooler 407 ⇒ air cooler 408 ⇒ exhaust port 480, and the compressed air generated in the compression section 401 is discharged from the exhaust port 480.

[0041] <Example of hardware configuration of computers (user site 201, production site 202, cloud site 203)> FIG. 5 is a block diagram showing an example of the hardware configuration of a computer. The computer 500 includes a processor 501, a storage device 502, an input device 503, an output device 504, and a communication interface (communication IF) 505. The processor 501, the storage device 502, the input device 503, the output device 504, and the communication IF 505 are connected via a bus 506. The processor 501 controls the computer 500. The storage device 502 serves as a working area for the processor 501. The storage device 502 is a non-transitory or temporary recording medium that stores various programs and data. Examples of the storage device 502 include a read-only memory (ROM), a random access memory (RAM), a hard disk drive (HDD), and a flash memory. The input device 503 inputs data. Examples of the input device 503 include a keyboard, a mouse, a touch panel, a numeric keypad, a scanner, a microphone, and a sensor. The output device 504 outputs data. The output device 504 may be, for example, a display, a printer, or a speaker. The communication IF 505 connects to a network and transmits and receives data.

[0042] <Data preprocessing example> Next, an example of data preprocessing by the data preprocessing unit 103 will be described.

[0043] 6 is an explanatory diagram showing an example of a sensor data table. The sensor data table 600 exists in the computer 500 that stores the sensor data 114. The sensor data table 600 is a table that has the sensor data 114 as entries, and includes fields such as date and time 601, discharge pressure 602, discharge temperature 603, ambient temperature 604, load factor 605, current value 606, power ON / OFF 607, and operating status 608.

[0044] Date and time 601 is the date and time when the sampling processing unit 102 sampled the analog data from the sensor 112. Discharge pressure 602 is the oil discharge pressure value at the date and time when the sampling processing unit 102 sampled the analog data from the pressure gauge 451. Discharge temperature 603 is the oil discharge temperature at the date and time when the sampling processing unit 102 sampled the analog data from the discharge thermometer 452. Ambient temperature 604 is the ambient temperature of the cooling target device 101 at the date and time when the sampling processing unit 102 sampled the analog data from the ambient thermometer 462.

[0045] The load factor 605 is a value indicating the proportion of the operating load on the motor, which is the heat source 111, at the date and time when the sampling processing unit 102 samples the analog data (frequency of the AC voltage) from the inverter 400, and increases or decreases depending on the frequency of the AC voltage converted by the inverter 400.

[0046] The current value 606 is the value of the current applied to the heat source 111 at the date and time when the sampling processing unit 102 sampled the analog data from the ammeter 411. The power ON / OFF 607 is a value indicating whether the power of the cooling target device 101 was ON or OFF at the date and time when the sampling processing unit 102 sampled the analog data from the sensor 112. The operating state 608 is a value indicating whether the cooling target device 101 was operating or idle at the date and time when the sampling processing unit 102 sampled the analog data from the sensor 112.

[0047] For example, if the load factor 605 of the sensor data 114 is equal to or greater than a threshold, the data preprocessing unit 103 outputs it as the feature amount 115, and if it is less than the threshold, it does not output it as the feature amount 115.

[0048] Furthermore, the data preprocessing unit 103 may output the sensor data 114 that is equal to or greater than a first threshold as the first feature amount 115, and may output the sensor data 114 that is less than a second threshold that is lower than the first threshold as the second feature amount 115. In this case, the construction unit 104 may construct the first sign detection model 117 using the first feature amount 115, or may construct the second sign detection model 117 using the second feature amount 115.

[0049] Furthermore, the data preprocessing unit 103 calculates a moving average value for a predetermined time width for each date and time 601 for the time-series data of the oil discharge temperature 603 from among the set of time-series sensor data 114. Then, the data preprocessing unit 103 may determine and output the sensor data 114 whose moving average value is in the top p% or more within the range between the maximum and minimum values ​​of the moving average value calculated for each date and time 601 as the feature amount 115, and may not output the sensor data 114 whose moving average value is less than the top p% as the feature amount 115.

[0050] Furthermore, the data preprocessing unit 103 may output the sensor data 114 at a certain date and time 601 (assumed to be t1) and statistics of the sensor data 114 from date and time t0 (a date and time preceding date and time t1 by a predetermined time T) to date and time t1 as the feature quantity 115 at date and time t1. The statistics may be, for example, at least one of the maximum value, minimum value, average value, variance, standard deviation, autocovariance, and autocorrelation of each element (discharge pressure 602, discharge temperature 603, ambient temperature 604, load factor 605, current value 606, power on / off 607, and operating state 608) included in the sensor data 114.

[0051] For example, assume that the sampling period of the sampling processing unit 102 is 30 minutes. If date and time t1 is 12:30 on a certain date and time, and the predetermined time T is 12 hours, then date and time t0 is 0:30 on that certain date and time. In this case, the data preprocessing unit 103 calculates statistics of the sensor data 114 for every 30 minutes from 0:30 (date and time t0) to 12:30 (date and time t1). The data preprocessing unit 103 outputs the sensor data 114 for 12:30 (date and time t1) and the statistics of the sensor data 114 for every 30 minutes from 0:30 (date and time t0) to 12:00, which is the date and time immediately before 12:30 (date and time t1), as the feature amount 115 for 12:30 (date and time t1).

[0052] If the sampling period is equal to or shorter than a predetermined period, the amount of sensor data 114 will become enormous. In particular, in the case of system configuration example 2 shown in Fig. 3, if the amount of data transmitted from the first communication control unit 210 to the third communication control unit 230 increases and there is a limit to the amount of data transmitted between the user site 201 and the cloud site 203, data transmission from the user site 201 to the cloud site 203 will become impossible. To prepare for such a case, the data preprocessing unit 103 converts the sensor data 114 into frequency components by fast Fourier transform.

[0053] For example, assume that the sampling period of the sampling processing unit 102 is 10 msec. If date and time t1 is 12:30 on a certain date and time, and the predetermined time T is 30 minutes, then date and time t0 is 12:00 on that certain date and time. In this case, the data preprocessing unit 103 converts the sensor data 114 obtained every 10 msec from 12:00 (date and time t0) to 12:30 (date and time t1) into frequency components by fast Fourier transform, and outputs the frequency components as feature quantities 115 for 12:30 (date and time t1). Note that the frequency component feature quantities 115 may be used directly at the production site 202 to build the sign detection model 117, or may be converted into time-series feature quantities 115 by performing an inverse fast Fourier transform at the production site 202.

[0054] <Example of building a predictive detection model> Next, we will explain an example of constructing a sign detection model by the constructing unit 104. The constructing unit 104 generates a training dataset (features 115 and positive / negative labels 116) and generates a sign detection model 117. First, we will explain how to generate the training dataset (features 115 and positive / negative labels 116).

[0055] 7 is an explanatory diagram showing an example of a training dataset table. The training dataset table 700 exists in the computer 500 that stores the features 115. The training dataset table 700 is a table with entries of the features 115 and the positive / negative labels 116, and includes fields such as date and time 601, discharge pressure 602, discharge temperature 603, ambient temperature 604, load factor 605, current value 606, power ON / OFF 607, operating status 608, and positive / negative label 116.

[0056] The construction unit 104 receives input of the date and time when an abnormality in cooling performance occurred through operation input from the operator of the production site 202. If the date and time 601 when the abnormality occurred is t1, the construction unit 104 sets the period from a date and time (t1-T) that is a predetermined time T before the date and time t1 when the abnormality occurred to date and time t1 as a positive period, and sets the positive / negative label 116 of the feature 115 for the positive period to "1", indicating positive. Furthermore, the construction unit 104 sets the period before the date and time (t1-T) to which the positive / negative label 116 is not assigned as a negative period, and sets the positive / negative label 116 of the feature 115 for the negative period to "0", indicating negative. In this way, a training dataset is generated for each feature 115.

[0057] The constructing unit 104 may also generate a training data set for each feature 115 by identifying the temperature rise period from the start to the end of the rise in the discharge temperature 603 in the set of time-series sensor data 114.

[0058] FIG. 8 is a graph showing time-series data of the discharge temperature 603 and the ambient temperature 604. In the graph 800, the construction unit 104 identifies a period of increasing trend in which the discharge temperature 603 continuously increases at a gradient equal to or greater than a predetermined value. The start date and time of the period is the date and time when the discharge temperature 603 reaches its lowest value, which is the date and time when the increase begins. Furthermore, if the discharge temperature 603 at a certain date and time drops by more than a predetermined temperature at the next date and time (for example, to a value equal to or less than the discharge temperature 603 at the date and time when the increase begins), the date and time when the increase ends is identified as the temperature increase period. The construction unit 104 identifies the period from the date and time when the increase begins to the date and time when the increase ends as the temperature increase period.

[0059] If the temperature rise period includes the date and time when the cooling performance abnormality occurred, the construction unit 104 sets the temperature rise period to a positive period and sets the positive / negative label 116 of the feature 115 for the positive period to "1", indicating positive. On the other hand, if the temperature rise period does not include the date and time when the cooling performance abnormality occurred, the construction unit 104 sets the temperature rise period to a negative period and sets the positive / negative label 116 of the feature 115 for the negative period to "0", indicating negative. The construction unit 104 may also set periods outside the temperature rise period to negative periods and set the positive / negative label 116 of the feature 115 for the negative period to "0", indicating negative.

[0060] Next, a description will be given of the generation of the sign detection model 117. The construction unit 104 uses a training dataset to generate the sign detection model 117, for example, by a decision tree, a random forest, or deep learning.

[0061] FIG. 9 is an explanatory diagram showing generation example 1 of the sign detection model 117. In generation example 1, a decision tree DT is generated. The decision tree DT has branching conditions for each element of the feature 115 for each node (for simplicity of explanation, in FIG. 9, discharge temperature 603, ambient temperature 604, and load factor 605 are used as examples). The construction unit 104 assigns the feature 115 to the decision tree DT, associates the number of normal cases (positive / negative label 116 is "0") with the number of abnormal cases (positive / negative label 116 is "1") for each terminal node, and calculates the accuracy of an abnormality occurrence. The accuracy of an abnormality is calculated for each terminal node by dividing the number of normal cases by (number of normal cases + number of abnormal cases). In this way, the decision tree DT is constructed.

[0062] Using this decision tree DT as a sign detection model 117, the sign detection unit 105 inputs a feature 115 to be predicted into the decision tree DT, thereby identifying the terminal node reached by the feature 115 to be predicted. If the sign accuracy of the reached terminal node is greater than a preset threshold, the diagnosis result for the input feature 115 to be predicted is positive (sign period), and if the sign accuracy of the reached terminal node is equal to or less than the preset threshold, the diagnosis result for the input feature 115 is negative (normal period).

[0063] For example, when the input feature 115 to be predicted reaches the terminal node 900, if the prognosis accuracy of 45% is greater than a preset threshold, the diagnostic result for the input feature 115 to be predicted is positive (prognosis), and if it is equal to or less than the preset threshold, the diagnostic result for the input feature 115 is negative (normal).

[0064] 10 is an explanatory diagram showing generation example 2 of the sign detection model 117. In generation example 2, a random forest RF is generated by combining multiple decision trees DT1, DT2, ..., DT50. Each of the decision trees DT1, DT2, ..., DT50 is constructed by combining elements that are the same as or different from the elements of the feature quantities 115 employed at the nodes of the other decision trees. Even when the same elements are used, the branching conditions may be different.

[0065] Using this random forest RF as the sign detection model 117, the sign detection unit 105 inputs the feature 115 to be predicted into the random forest RF, thereby identifying the terminal node reached by the feature 115 to be predicted for each of the decision trees DT1, DT2, ..., DT50. The sign detection unit 105 takes a majority vote of the diagnosis results (normal period or sign period) for each of the decision trees DT1, DT2, ..., DT50, and outputs the final diagnosis result of the feature 115 to be predicted. In the case of FIG. 10, the majority vote determines the sign period.

[0066] As described above, according to this embodiment, by focusing on the load on the device to be cooled 101 or the oil temperature rise, it is possible to detect a sign of an abnormality occurring in the cooling performance of the oil cooling unit 113. Therefore, when focusing on the load on the device to be cooled 101, it is possible to detect a sign of an abnormality occurring in the cooling performance of the oil cooling unit 113 without relying on the oil temperature rise. Furthermore, when focusing on the oil temperature rise, it is possible to detect a sign of an abnormality occurring directly due to the oil temperature rise. Therefore, in either case, it is possible to prevent performance degradation, stoppage of the device to be cooled 101, or failure of the device to be cooled 101 due to the occurrence of an abnormality in the device to be cooled 101.

[0067] In the above-described embodiment, an air compressor has been taken as an example of the apparatus to be cooled 101, but the apparatus to be cooled 101 may also be a rolling mill or an engine.

[0068] The present invention is not limited to the above-described embodiments, and includes various modifications and equivalent configurations within the spirit and scope of the appended claims. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to configurations including all of the described configurations. Furthermore, part of the configuration of one embodiment may be replaced with the configuration of another embodiment. Furthermore, the configuration of another embodiment may be added to the configuration of one embodiment. Furthermore, part of the configuration of each embodiment may be added to, deleted from, or replaced with other configurations.

[0069] Furthermore, the aforementioned configurations, functions, processing units, processing means, etc. may be realized in part or in whole in hardware, for example by designing them as integrated circuits, or may be realized in software by having a processor interpret and execute a program that realizes each function.

[0070] Information such as programs, tables, files, etc. that realize each function can be stored in storage devices such as memory, hard disks, SSDs (Solid State Drives), or recording media such as IC (Integrated Circuit) cards, SD cards, and DVDs (Digital Versatile Discs).

[0071] In addition, the control lines and information lines shown are those that are considered necessary for explanation, and do not necessarily represent all the control lines and information lines that are necessary for implementation. In reality, it can be assumed that almost all components are interconnected. [Explanation of symbols]

[0072] 100 Predictive Detection System 101 Cooling target equipment 102 Sampling processing section 103 Data preprocessing section 104 Construction Department 105 Predictive detection unit 111 Heat source 112 Sensors 113 Oil cooling section 114 Sensor Data 115 Features 116 Positive and Negative Labels 117 Predictive Detection Model 118 Diagnosis Results 201 User Site 202 Operational Site 203 Cloud Site 210 First communication control unit 220 Second communication control unit 230 Third Communication Control Unit

Claims

1. an acquisition unit that acquires time-series data indicating an operating status of a cooling target device having a heat source and a cooling unit that cools the heat source with a cooling medium and including a temperature of the cooling medium; a sign detection unit that performs sign detection of deterioration in cooling performance of the cooling unit using the time-series data acquired by the acquisition unit and a sign detection model, The symptom detection model is constructed using time-series feature quantities of specific time-series data in which the temperature of the cooling medium is equal to or higher than a threshold value among the time-series data, and a label indicating whether the cooling performance of the cooling unit is normal or abnormal. A symptom detection device characterized by:

2. The symptom detection device according to claim 1, The symptom detection model includes a first symptom detection model constructed using a first feature amount, which is time-series data in which the temperature of the cooling medium is equal to or greater than a first threshold value, among the time-series data, and a second symptom detection model constructed using a second feature amount, which is time-series data in which the temperature of the cooling medium is less than a second threshold value that is lower than the first threshold value, among the time-series data. A symptom detection device characterized by:

3. an acquisition unit that acquires time-series data indicating an operating status of a cooling target device having a heat source and a cooling unit that cools the heat source with a cooling medium and including a temperature of the cooling medium; a preprocessing unit that calculates, as a time-series feature, specific time-series data in which the temperature of the cooling medium is equal to or higher than a threshold value from among the time-series data acquired by the acquisition unit; a sign detection unit that performs sign detection of deterioration in cooling performance of the cooling unit using the time-series feature amount calculated by the preprocessing unit and a sign detection model; A sign detection device comprising:

4. The symptom detection device according to claim 3, The sign detection model is constructed using time-series feature quantities of specific time-series data in which the temperature of the cooling medium is equal to or higher than a threshold value among the time-series data, and a label indicating whether the cooling performance of the cooling unit is normal or abnormal. A symptom detection device characterized by:

5. The symptom detection device according to claim 4, the sign detection model includes a first sign detection model constructed using a first feature amount which is time-series data in which the temperature of the cooling medium is equal to or greater than a first threshold value among the time-series data, and a second sign detection model constructed using a second feature amount which is time-series data in which the temperature of the cooling medium is less than a second threshold value which is lower than the first threshold value among the time-series data, the preprocessing unit calculates, as a first feature, specific time-series sensor data among the time-series sensor data in which the temperature of the cooling medium is equal to or greater than a first threshold, and calculates, as a second feature, specific time-series sensor data among the time-series sensor data in which the temperature of the cooling medium is less than a second threshold that is lower than the first threshold, the sign detection unit performs sign detection using a first sign detection model when the feature amount calculated by the preprocessing unit is a first feature amount, and performs sign detection using a second sign detection model when the feature amount acquired by the preprocessing unit is a second feature amount; A symptom detection device characterized by:

6. The symptom detection device according to claim 1 or 3, The sign detection model calculates a sign accuracy using the number of normal cases and the number of abnormal cases in at least two or more combinations of data representing the operating state of the cooling target devices, and determines that a sign of abnormality is present when the calculated sign accuracy exceeds a predetermined threshold value set for each of the combinations. A symptom detection device characterized by:

7. The symptom detection device according to claim 6, the device to be cooled is an air compressor, The data representing the operating state includes any one of a discharge temperature, an ambient temperature, and a load factor. A symptom detection device characterized by:

8. A symptom detection method executed by a symptom detection device having a processor that executes a program and a storage device that stores the program, The processor: an acquisition process for acquiring time-series data indicating an operating status of a cooling target device having a heat source and a cooling unit that cools the heat source with a cooling medium, and including a temperature of the cooling medium; a sign detection process for detecting a sign of deterioration in cooling performance of the cooling unit using the time-series data acquired in the acquisition process and a sign detection model; In the precursor detection process, the processor performs precursor detection using the precursor detection model constructed using time-series features of specific time-series data in which the temperature of the cooling medium is above a threshold value among the time-series data, and a label indicating whether the cooling performance of the cooling unit is normal or abnormal.

9. A symptom detection method executed by a symptom detection device having a processor that executes a program and a storage device that stores the program, The processor: an acquisition process for acquiring time-series sensor data indicating an operating status of a cooling target device having a heat source and a cooling unit that cools the heat source with a cooling medium and including a temperature of the cooling medium; a pre-processing step of calculating specific time-series sensor data in which the temperature of the cooling medium is equal to or higher than a threshold value as a time-series feature quantity from among the time-series sensor data; a sign detection process for performing sign detection using the time-series feature amount acquired by the preprocessing and a sign detection model; A sign detection method characterized by executing the above.

Images