State monitoring apparatus

Abstract

To provide a state monitoring apparatus capable of successively monitoring a state of an object conveyed on a conveyor and precisely detect an anomaly.SOLUTION: A state monitoring apparatus includes: a first information acquisition apparatus which acquires first information, which is information on an object, at an upstream position on a conveyor; a second information acquisition apparatus which acquires second information, which is information on the object, at a downstream position; a prediction information generation unit which predicts a normal state of the object at the downstream position, based on the first information, and applies convolution processing to predicted information to generate prediction information; an information processing unit which generates processed second information by applying convolution to the second information; and an anomaly determination unit which compares the prediction information with the processed second information to determine whether the processed second information includes an anomaly.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present disclosure relates to a condition monitoring device. [Background technology]

[0002] For example, in non-burnable waste shredding and processing facilities, conveyors are used to transport waste before and after the shredding equipment. The waste transported on these conveyors includes items that may generate heat or catch fire before or after the shredding process, such as lithium-ion batteries and gas cylinders. For fire safety reasons, such facilities require technology to monitor the conveyors with monitoring devices such as cameras and detect any abnormalities.

[0003] The following Patent Document 1 describes a method for detecting anomalies in input data by generating training data based on normal data and comparing the training data with input data acquired from an actual object. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 6740247 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the above method is limited to comparing input data with training data at a specific location, and therefore cannot adequately monitor and detect anomalies in objects whose positions change over time as they are transported on a conveyor, as described above.

[0006] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a condition monitoring device that can sequentially monitor the condition of objects transported on a conveyor and precisely detect abnormalities. [Means for solving the problem]

[0007] In order to solve the above problem, the condition monitoring device of the present disclosure is a condition monitoring device for monitoring the condition of an object transported on a conveyor, and includes: a first information acquisition device that acquires first information, which is information about the object, at an upstream position on the conveyor; a second information acquisition device that acquires second information, which is information about the object, at a position downstream of the first information acquisition device on the conveyor; a predicted information generation unit that predicts the state of the object at the second position under normal conditions based on the first information and generates predicted information by performing a convolution process on the predicted information; an information processing unit that generates processed second information by performing a convolution on the second information; and an abnormality determination unit that determines whether the processed second information contains an abnormality by comparing the predicted information with the processed second information. [Effects of the Invention]

[0008] According to the present disclosure, it is possible to provide a condition monitoring device that can sequentially monitor the condition of an object being transported on a conveyor and precisely detect abnormalities. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a schematic diagram illustrating a configuration of a state monitoring device according to an embodiment of the present disclosure. [Figure 2] FIG. 2 is a functional block diagram of a control device according to an embodiment of the present disclosure. [Figure 3] 10 is a flowchart illustrating a processing flow of a control device according to an embodiment of the present disclosure. [Figure 4] FIG. 10 is an explanatory diagram illustrating an example of a convolution process according to an embodiment of the present disclosure. [Figure 5] FIG. 1 is a hardware configuration diagram illustrating a configuration of a computer according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0010] A condition monitoring device 1 according to an embodiment of the present disclosure will be described below with reference to Figures 1 to 5. This condition monitoring device 1 is suitably used to monitor objects 100 on a transport conveyor 2 installed, for example, at a facility for crushing and processing non-burnable waste.

[0011] (Conveyor configuration) As shown in FIG. 1, the conveyor 2 has a plurality of rollers 21, a belt 22 stretched over these rollers 21, and a drive unit (not shown). The rollers 21 are driven to rotate by the drive unit, causing the belt 22 to move forward and backward. In the following description, it is assumed that the speed at which the belt 22 moves forward and backward is constant. Objects 100 (non-burnable waste, etc.) sent from other preceding devices are continuously placed on the belt 22. For simplicity of explanation, the following description will be based on an example in which one specific object 100 is placed. The side from which the objects 100 flow may be referred to as the upstream side, and the side from which they flow may be referred to as the downstream side.

[0012] (Configuration of the condition monitoring device) The status monitoring device 1 includes a first information acquisition device 11, a second information acquisition device 12, and a control device 13.

[0013] (First information acquisition device / Second information acquisition device) The first information acquisition device 11 is an infrared camera that captures the object 100 on the conveyor 2 at an upstream position. That is, it is capable of capturing images of the surface temperature and the deep temperature of the object 100 without contact and transmitting the images as electrical signals to the outside as "first information." The first information is two-dimensional image data having a spatial distribution.

[0014] The second information acquisition device 12 is disposed downstream of the first information acquisition device 11. The second information acquisition device 12 is an infrared camera that captures the object 100 transported on the conveyor 2 at a downstream position. Because the speed of the belt 22 is constant, by calculating the product of the speed and the transport time, the second information acquisition device 12 can capture the same object 100 as the object 100 imaged by the first information acquisition device 11 at a later time. The data captured by the second information acquisition device 12 is called "second information." The second information is two-dimensional image data having a spatial distribution. It is desirable that the angles of view of the first information acquisition device 11 and the second information acquisition device 12, i.e., the imaging range and angle relative to the object 100 and the conveyor 2, are the same.

[0015] (Control device configuration) Based on the first information and second information, the control device 13 detects whether or not there is an abnormality in the object 100 on the conveyor 2. Specifically, as shown in Fig. 2, the control device 13 has an information acquisition unit 31, a prediction information generation unit 32, an information processing unit 33, an abnormality determination unit 34, and a memory unit 35.

[0016] The information acquisition unit 31 acquires the first information and the second information from the first information acquisition device 11 and the second information acquisition device 12, respectively, and stores them in the storage unit .

[0017] The prediction information generating unit 32 predicts the state of the object 100 at a downstream position (i.e., the position of the second information acquisition device 12) under normal conditions based on the first information, and generates prediction information by performing convolution processing on the predicted information. Here, the objects 100 transported on the conveyor 2 may include objects that generate heat or ignite when subjected to a crushing process, such as lithium-ion batteries or gas cylinders. A state in which such a heat-generating or igniting object 100 is present is defined as an "abnormal state," and a state in which such an object 100 is not present is defined as a "normal state." The prediction information generating unit 32 assumes that the object 100 included in the first information will not undergo any state changes, such as ignition or heat generation, even when it reaches a downstream position under normal conditions. Based on this assumption, the prediction information generating unit 32 generates a prediction image of the object 100 when it reaches the downstream position based on a prediction model prepared in advance. The prediction information is generated by further performing convolution processing on this prediction image. The prediction model is, for example, a deep learning model (supervised learning model) such as a deep neural network (DNN).

[0018] In the above convolution process, the acquired two-dimensional image data is divided into multiple sections and abnormal data (outliers) are removed, as shown in Fig. 4. Finally, statistics (e.g., average temperature, upper limit, or lower limit) are derived for multiple adjacent sections (four sections in this case).

[0019] The information processing unit 33 performs the same convolution process as described above on the second information. This generates processed second information. The number of sections in the processed second information is controlled to be the same as that of the predicted information. In other words, the processed second information and the predicted information can be compared with each other.

[0020] The abnormality determination unit 34 determines whether the processed second information contains an abnormality by comparing the processed second information with the prediction information. That is, if the processed second information contains a section containing an abnormal value compared to the prediction information indicating a normal state, it is determined that an abnormality such as a fire or heat generation has occurred.

[0021] (Processing flow of the control device) Next, the processing flow of the control device 13 will be described with reference to FIG. 3. First, the information acquisition unit 31 acquires first information (step S1). Next, in step S2, the prediction information generation unit 32 generates prediction information. After that, the information acquisition unit 31 acquires second information about the object 100 after it has been transported on the conveyor 2 (step S3). Next, in step S4, the information processing unit 33 performs a convolution process on the second information to generate processed second information. In step S5, the abnormality determination unit 34 compares the prediction information with the processed second information to determine whether or not an abnormality exists. If no abnormality is detected in step S5, the process returns to steps S1 and S3. On the other hand, if an abnormality is determined to exist, measures to eliminate the abnormality are taken in step S6. For example, measures include removing the object 100 in which the abnormality has occurred from the conveyor 2 or spraying a fire extinguisher.

[0022] (Action and effect) Here, the waste transported on the conveyor 2 includes items that may generate heat or ignite before or after the crushing process, such as lithium-ion batteries and gas cylinders. For fire safety reasons, such facilities require technology to detect abnormalities while monitoring the conveyor 2 with a monitoring device such as a camera. However, conventional technology is limited to comparing input data with training data at a specific location. This has led to the problem that it is not possible to adequately monitor and detect abnormalities for objects whose position changes over time as they are transported on the conveyor 2. To solve this problem, the present embodiment employs the above-described configurations.

[0023] According to the above configuration, by comparing the processed second information regarding the actual state of the object 100 with the predicted information as normal data predicted and generated in advance, any abnormalities contained in the processed second information can be immediately detected. In other words, under normal conditions, the object 100 transported on the conveyor 2 moves from upstream to downstream without any changes such as ignition or explosion. On the other hand, if an abnormality such as ignition occurs during transport, the processed second information acquired downstream will contain the abnormality. By comparing and matching this processed second information with the predicted information, any abnormalities in the object 100 can be detected without fail. In addition, both the predicted information and the processed second information are obtained by performing convolution processing on the raw first information and second information. This reduces the data size handled, enabling the above-mentioned anomaly detection process to be performed more quickly and accurately. Therefore, the status of a series of objects 100 continuously flowing on the conveyor 2 can be continuously monitored in real time without stopping the movement of the conveyor 2. This improves the operational efficiency of treatment facilities and further enhances fire safety.

[0024] According to the above configuration, the convolution process divides the two-dimensional image data into multiple sections and acquires statistics for each section. This allows the data size to be compressed compared to the two-dimensional image data actually acquired, thereby speeding up processing. In addition, by performing the convolution process, it is possible to reduce misalignment and minute noise caused by the placement of the first information acquisition device 11 and the second information acquisition device 12, enabling more accurate anomaly detection.

[0025] According to the above configuration, the predicted information and the processed second information are compared between specific sections included in the mosaic image. This makes it possible to accurately identify which part of the first information obtained upstream has an abnormality. Therefore, if an abnormality is found, fire prevention measures or removal processes can be quickly carried out on the part or object 100 where the abnormality occurred. As a result, it is possible to further improve the safety and operational stability of the facility.

[0026] According to the above configuration, an infrared camera is included as the first information acquisition device 11 and the second information acquisition device 12. This makes it possible to accurately capture changes in the temperature of the object 100 in a non-contact manner. Therefore, an abnormally high temperature state caused by a fire or the like can be quickly and accurately detected.

[0027] (Other embodiments) The above describes in detail the embodiments of the present disclosure with reference to the drawings, but the specific configuration is not limited to this embodiment, and design changes and the like are also included within the scope that does not deviate from the gist of the present disclosure.

[0028] For example, the first information acquisition device 11 and the second information acquisition device 12 may include a concentration sensor capable of detecting the concentration of a specific chemical species. That is, it is possible to adopt a configuration in which the above-mentioned infrared camera and concentration sensor are provided side by side at positions upstream and downstream.

[0029] According to this configuration, concentration sensors are included as the first information acquisition device 11 and the second information acquisition device 12. This makes it possible to accurately capture, in a non-contact manner, fires, temperature rises, and the like in internal areas that cannot be seen with the naked eye. Therefore, abnormally high temperatures caused by fires and the like can be quickly and accurately detected.

[0030] In addition to the infrared camera and density sensor, a visible light camera for capturing visible light images may be provided.

[0031] Note that the control device 13 in the embodiment of the present disclosure may change the order of processing as long as appropriate processing is performed.

[0032] The storage unit 35 and other storage devices in the embodiments of the present disclosure may be provided anywhere within the range where appropriate information can be transmitted and received. Furthermore, there may be multiple storage units 35 and other storage devices, and data may be stored in a distributed manner within the range where appropriate information can be transmitted and received.

[0033] The above-described processing steps performed by the control device 13 are stored in the form of a program on a recording medium that can be read by the computer 200, and the above processing is performed by the computer 200 reading and executing this program. A specific example of the computer 200 is shown below.

[0034] As shown in FIG. 5, the computer 200 includes a CPU 201, a main memory 202, a storage 203, and an interface 204. For example, the above-described control device 13 is implemented in a computer 200. The operations of the above-described processing units are stored in the form of a program in a storage 203. The CPU 201 reads the program from the storage 203, loads it into the main memory 202, and executes the above-described processing in accordance with the program. The CPU 201 also allocates a storage area in the main memory 202 corresponding to the above-described storage unit 35 in accordance with the program.

[0035] Examples of storage 203 include a hard disk drive (HDD), a solid state drive (SSD), a magnetic disk, a magneto-optical disk, a compact disc read only memory (CD-ROM), a digital versatile disc read only memory (DVD-ROM), and a semiconductor memory. Storage 203 may be an internal medium directly connected to the bus of computer 200, or an external medium connected to computer 200 via interface 204 or a communication line. Furthermore, when this program is distributed to computer 200 via a communication line, computer 200 that receives the program may load the program into main memory 202 and execute the above-mentioned processing. Storage 203 is a non-transitory tangible storage medium.

[0036] The program may also implement some of the functions described above. Furthermore, the program may be a file that can implement the functions described above in combination with a program already recorded in computer 200, a so-called differential file (differential program).

[0037] In addition to or instead of the above configuration, a custom LSI (Large Scale Integrated Circuit) such as a PLD (Programmable Logic Device), an ASIC (Application Specific Integrated Circuit), a GPU (Graphics Processing Unit), or similar processing devices may be provided. Examples of PLDs include PAL (Programmable Array Logic), GAL (Generic Array Logic), CPLD (Complex Programmable Logic Device), and FPGA (Field Programmable Gate Array). In this case, some or all of the functions realized by the processor may be realized by the integrated circuit.

[0038] <Additional Notes> The state monitoring device 1 described in each embodiment can be understood, for example, as follows.

[0039] (1) The condition monitoring device 1 of the first aspect is a condition monitoring device 1 for monitoring the condition of an object 100 transported on a conveyor 2, and includes a first information acquisition device 11 that acquires first information, which is information about the object 100, at an upstream position on the conveyor 2; a second information acquisition device 12 that acquires second information, which is information about the object 100, at a position downstream of the first information acquisition device 11 on the conveyor 2; a predicted information generation unit 32 that predicts the state of the object 100 at the downstream position under normal conditions based on the first information and generates predicted information by performing a convolution process on the predicted information; an information processing unit 33 that generates processed second information by performing a convolution process on the second information; and an abnormality determination unit 34 that compares the predicted information with the processed second information to determine whether the processed second information contains an abnormality.

[0040] According to the above configuration, by comparing the processed second information regarding the actual state of the object 100 with predicted information as normal data that has been predicted and generated in advance, it is possible to immediately detect abnormalities contained in the processed second information.

[0041] (2) The condition monitoring device 1 according to the second aspect is the condition monitoring device 1 of (1), wherein the first information and the second information are two-dimensional image data relating to the object 100, and the prediction information generation unit 32 and the information processing unit 33 divide the two-dimensional image data into a plurality of sections, acquire statistics for each section, and generate the prediction information as a mosaic image or the processed second information.

[0042] According to the above configuration, the data size can be compressed compared to the two-dimensional image data that is actually acquired, thereby enabling faster processing.

[0043] (3) The condition monitoring device 1 according to the third aspect is the condition monitoring device 1 of (2), in which the abnormality determination unit 34 determines whether or not the abnormality is present by comparing the specific section included in the mosaic image as the predicted information with the specific section included in the mosaic image as the processed second information.

[0044] According to the above configuration, the predicted information is compared with the processed second information between specific sections included in the mosaic image, which makes it possible to accurately identify the location of the abnormality in the first information obtained upstream.

[0045] (4) The condition monitoring device 1 according to the fourth aspect is a condition monitoring device 1 according to any one of aspects (1) to (3), in which the first information acquisition device 11 and the second information acquisition device 12 include an infrared camera.

[0046] According to the above configuration, it is possible to accurately capture the change in temperature of the object 100 in a non-contact state.

[0047] (5) The condition monitoring device 1 according to the fifth aspect is a condition monitoring device 1 according to any one of aspects (1) to (4), wherein the first information acquisition device 11 and the second information acquisition device 12 include a concentration sensor capable of detecting the concentration of a specific chemical species.

[0048] According to the above configuration, it is possible to accurately detect ignition, temperature rise, etc. in an internal area that cannot be visually confirmed in a non-contact manner. [Explanation of symbols]

[0049] REFERENCE SIGNS LIST 1... Condition monitoring device 2... Conveyor 11... First information acquisition device 12... Second information acquisition device 13... Control device 21... Roller 22... Belt 31... Information acquisition unit 32... Prediction information generation unit 33... Information processing unit 34... Abnormality determination unit 35... Memory unit 100... Object 200... Computer 201... CPU 202... Main memory 203... Storage 204... Interface

Claims

1. A condition monitoring device for monitoring the condition of an object transported on a conveyor, comprising: a first information acquisition device that acquires first information, which is information about the object, at an upstream position on the conveyor; a second information acquisition device that acquires second information, which is information about the object, at a position downstream of the first information acquisition device on the conveyor; a prediction information generating unit that predicts a state of the object at the downstream position under normal conditions based on the first information, and generates prediction information by performing a convolution process on the predicted information; an information processing unit that generates processed second information by performing convolution on the second information; an abnormality determination unit that determines whether or not the processed second information includes an abnormality by comparing the prediction information with the processed second information; A condition monitoring device comprising:

2. the first information and the second information are two-dimensional image data relating to the object; 2. The condition monitoring device according to claim 1, wherein the prediction information generation unit and the information processing unit divide the two-dimensional image data into a plurality of sections, acquire statistics for each section, and generate the prediction information or the processed second information as a mosaic image.

3. The condition monitoring device according to claim 2, wherein the abnormality determination unit determines whether or not the abnormality is present by comparing the specific sections included in the mosaic image as the prediction information with the specific sections included in the mosaic image as the processed second information.

4. The condition monitoring device according to claim 1 , wherein the first information acquisition device and the second information acquisition device include an infrared camera.

5. The condition monitoring device according to claim 1 , wherein the first information acquisition device and the second information acquisition device include a concentration sensor capable of detecting a concentration of a specific chemical species.

Images