Information processing device, operation support system, information processing method, and information processing program

Abstract

A versatile device that can be used for assisting the operation of a machine is provided. The information processing device (1A) comprises: a data acquisition unit (101) that acquires sensor data from a plurality of sensors installed in the machine; a relationship index calculation unit (103) that calculates a relationship index representing the correlation of the group of sensors based on the distribution of each sensor data acquired from the group of sensors; and a map generation unit (104) that generates a map representing the state of the machine using the relationship index.

Description

Technical Field

[0001] The present invention relates to an information processing device and the like that generate information that can be used to assist in the operation of a machine. Background Art

[0002] Operators who operate mechanical equipment understand the state of the mechanical equipment based on output values ​​from sensors installed on the mechanical equipment and operate the mechanical equipment. Systems that assist in the operation of such mechanical equipment have been developed. For example, Patent Document 1 below discloses a plant support device that (1) outputs estimated values ​​of the plant based on process data containing measurement data such as flow rate, pressure, and temperature, and (2) predicts the future state of the plant based on the estimated values ​​and outputs the predicted values ​​of the plant.

[0003] Prior art literature

[0004] Patent Literature

[0005] Patent Document 1: Japanese Patent Publication No. 2018-112903 Summary of the Invention

[0006] (1) Technical issues to be solved

[0007] Conventional technologies such as those described above use simulation models based on physical / chemical model formulas to estimate factory values. These models require multiple parameter settings, resulting in high production costs. Furthermore, these models suffer from a technical limitation: their factory-specific nature limits their versatility.

[0008] An object of one embodiment of the present invention is to realize an information processing device or the like that can generate information that can be used to assist in the operation of a machine and that can be easily generalized.

[0009] (2) Technical solution

[0010] In order to solve the above-mentioned technical problems, an information processing device of one embodiment of the present invention comprises: a data acquisition unit, which acquires sensor data, wherein the sensor data is the output value of a plurality of sensors provided in a mechanical device during a specified period, or a numerical value calculated using the output value; a relationship index calculation unit, which calculates a relationship index representing the correlation of the group based on the distribution status of each sensor data acquired from a sensor of a group of sensors consisting of two of the plurality of the above-mentioned sensors; and a status information generation unit, which uses the above-mentioned relationship index calculated for each group of the plurality of the above-mentioned sensors to generate information representing the status of the above-mentioned mechanical device during the above-mentioned specified period.

[0011] In addition, in order to solve the above-mentioned technical problems, an operation assistance system of one embodiment of the present invention assists the operation of mechanical equipment, and the operation assistance system includes: an information processing device; and multiple sensors, which are arranged on the above-mentioned mechanical equipment; the above-mentioned information processing device obtains sensor data, and the sensor data is the output value of the above-mentioned multiple sensors during a specified period, or a numerical value calculated using the output value, namely sensor data, and the above-mentioned information processing device calculates a relationship index representing the correlation of the group based on the distribution status of each sensor data obtained from the sensor of a group of sensors consisting of two of the multiple sensors, and uses the above-mentioned relationship index calculated for each group of the multiple sensors to generate information representing the state of the above-mentioned mechanical equipment during the above-mentioned specified period and outputs it.

[0012] In addition, in order to solve the above-mentioned technical problems, an information processing method of one embodiment of the present invention is executed by an information processing device, and the information processing method includes: a data acquisition step of acquiring sensor data, wherein the sensor data is the output value of multiple sensors provided in the mechanical equipment during a specified period, or a numerical value calculated using the output value; a relationship index calculation step of calculating a relationship index representing the correlation of the group based on the distribution status of each sensor data acquired from the sensor of a group of sensors consisting of two of the multiple sensors; and a status information generation step of using the above-mentioned relationship index calculated for each group of the multiple sensors to generate information representing the status of the above-mentioned mechanical equipment during the above-mentioned specified period.

[0013] (3) Beneficial effects

[0014] According to one embodiment of the present invention, information indicating the state of a machine, useful for assisting the operation of the machine, can be generated. Furthermore, according to one embodiment of the present invention, information indicating the state of another machine can be generated by acquiring sensor data from that machine, thus facilitating generalization. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a block diagram showing an example of the configuration of main parts of the information processing device according to the first embodiment of the present invention.

[0016] Figure 2 This is a diagram for explaining the outline of an operation support system including the above-mentioned information processing device.

[0017] Figure 3 This is a diagram showing an example of dividing sensor data and an example of calculating a relationship index.

[0018] Figure 4 1 is a diagram showing an example of the division result of the values ​​of sensor data outputted by two sensors.

[0019] Figure 5An example of a correlation map representing the status of a waste incineration plant is shown.

[0020] Figure 6 This is a diagram showing an example of weight setting.

[0021] Figure 7 This is a diagram showing an example of changes in the correlation map.

[0022] Figure 8 This is a flowchart showing an example of processing executed by the above-mentioned information processing device.

[0023] Figure 9 This is a diagram showing the configuration of an information processing device according to Embodiment 2 of the present invention and the flow of processing executed by the information processing device.

[0024] Figure 10 This is a diagram showing an example of calculating the stability index.

[0025] Figure 11 This is a diagram showing the configuration of an information processing device and an overview of a method for predicting sensor data values ​​according to a third embodiment of the present invention.

[0026] Figure 12 This is a block diagram showing a configuration example of an information processing device according to a fourth embodiment of the present invention. DETAILED DESCRIPTION

[0027] (Implementation 1)

[0028] (System Overview)

[0029] based on Figure 2 An overview of a driving support system 100 according to an embodiment of the present invention will be described. Figure 2 This diagram is used to describe the outline of the operation support system 100. The operation support system 100 is a system for supporting the operation of various machines and includes sensors S1 to Sn and an information processing device 1A as shown. When it is not necessary to distinguish between sensors S1 to Sn, they are simply referred to as sensor S.

[0030] In this embodiment, an example of using the operation support system 100 to assist the operation of a waste incineration plant P will be described. Waste incineration plant P includes an incinerator that burns waste and a power generation facility that generates electricity using the heat generated in the incinerator. Furthermore, the operation support system 100 can assist any mechanical device, not limited to the waste incineration plant P, as long as its operating status can be monitored using sensors and other means and its operation can be controlled (operated) manually or automatically. For example, the operation support system 100 can also assist the operation of wind turbines and other equipment.

[0031] Sensors S detect predetermined physical quantities related to the plant status or their changes, and output sensor data representing the detection results to the information processing device 1A. Sensors S1-Sn each detect different objects. As described above, in this embodiment, sensors S are installed at various locations within the waste incineration plant P to assist in its operation. For example, sensors S may include a temperature sensor that indicates the temperature within the incinerator, a sensor that detects the carbon monoxide concentration in the exhaust gas, and the like.

[0032] In addition, the sensor S can be set for a part of the mechanical equipment of the waste incineration plant P. For example, if a plurality of sensors S are set for the incinerator, the information processing device 1A can output information indicating the state of the incinerator. More specifically, in order to output information indicating the combustion state, a thermometer, an air flow meter, a grate speed measuring device, etc., which are arranged around the incinerator, can also be used. In addition, if a plurality of sensors S are set for a part of the mechanical equipment, the information processing device 1A can output information related to the part. For example, if a plurality of sensors S related to the exhaust gas from the incinerator (for example, a thermometer in the furnace, a measuring device for the thickness of the garbage layer, a CO concentration meter, etc.) are set, the information processing device 1A can output information related to the exhaust gas.

[0033] As described in detail below, the information processing device 1A obtains the output values ​​of multiple sensors S during a specified period, that is, sensor data. Next, the information processing device 1A calculates a relationship index representing the correlation of the group based on the distribution status of each sensor data obtained from the sensor S of the group of sensor S consisting of two of the multiple sensors S (how each sensor data is distributed). Then, the information processing device 1A uses the relationship index calculated for each group of multiple sensors S to generate and output information representing the state of the mechanical equipment during the above-mentioned specified period. In this way, information representing the state of the mechanical equipment, which is useful for assisting the operation of the mechanical equipment, can be generated. In addition, if the information processing device 1A obtains sensor data from other mechanical equipment, it can generate information representing the state of the mechanical equipment, so it is easy to generalize.

[0034] Furthermore, the inventors of this application have experimentally confirmed that the relationship between the distribution patterns of sensor data within a group of sensors S reflects the state of the mechanical equipment. Furthermore, the above-described structure has also confirmed that it is possible to understand nonlinear correlations, which are difficult to handle using conventional analysis methods such as those described in Patent Document 1.

[0035] exist Figure 2In the example, the information indicating the state of the mechanical equipment is mappings M111 to M113 indicating the state of the waste incineration plant P. Mapping M111 is generated using sensor data during a stable period when the waste incineration plant P is in a stable operating state. Mapping M113 is generated using sensor data during an unstable period when the operating state of the waste incineration plant P is unstable. In addition, mapping M112 is generated using sensor data during the period immediately before the unstable period, that is, during the unstable period immediately before the unstable period. In addition, the standard of stable to unstable can be set arbitrarily. For example, the period when the amount of steam used for power generation is within the normal range can be set as the stable period, and the period when the amount of steam exceeds the normal range can be set as the unstable period.

[0036] As shown in the figure, maps M111 to M113 reflect the state of the waste incineration plant P from the stable period to the unstable period. More specifically, maps M111 to M112 show that the overall color of the map becomes lighter from the stable period to the period immediately before the unstable period. Furthermore, maps M112 to M113 show that the overall color of the map becomes even lighter from the period immediately before the unstable period to the unstable period.

[0037] Therefore, the operator of the waste incineration plant P can determine whether the state of the waste incineration plant P is approaching either a stable period or an unstable period based on the color density and color distribution of the overall map. In other words, if the overall color density and color distribution of the map generated based on the latest sensor data are close to map M111, the operator can determine that the state is stable. If they are close to map M113, the operator can determine that the state is unstable.

[0038] Furthermore, the map M112 of the period immediately preceding instability and the map M111 of the stable period can be distinguished by appearance. In other words, the map generated by the information processing device 1A shows signs of instability. Therefore, by referring to the map generated by the information processing device 1A, operators and others can take measures to stabilize the waste incineration plant P before the instability period. This can prevent the waste incineration plant P from becoming unstable.

[0039] (Device Structure)

[0040] based on Figure 1 A more detailed configuration of the information processing device 1A will be described. Figure 1This is a block diagram showing an example of the main component configuration of an information processing device 1A. As shown, the information processing device 1A includes a control unit 10A that centrally controls the various components of the information processing device 1A, and a storage unit 11 that stores various data used by the information processing device 1A. The control unit 10A also includes a data acquisition unit 101, a partitioning unit 102, a relationship index calculation unit 103, and a map generation unit 104.

[0041] The information processing device 1A further includes an input unit 12 for receiving information input to the information processing device 1A, and an output unit 13 for outputting information from the information processing device 1A. The input unit 12 and the output unit 13 may also be external devices to the information processing device 1A. In the example described in this embodiment, the input unit 12 is an input interface unit that receives sensor data output by a sensor S, and the output unit 13 is a display device that displays output images. The input unit 12 and the output unit 13 are not limited to these examples, as long as they have information input and output functions.

[0042] The data acquisition unit 101 acquires the output value of the above-mentioned sensor S, that is, sensor data. In addition, the data acquisition unit 101 can also acquire a numerical value calculated using the output value of the sensor S (for example, a value after normalizing the output value, a value after removing noise components from the output value, etc.) as sensor data. Sensor data can also be called process data. In addition, in addition to sensor data, the data acquisition unit 101 can also acquire the action setting values ​​of various mechanical equipment included in the waste incineration plant P (for example, the setting value of the action speed, etc.). Such action setting values ​​can also be processed in the same way as sensor data.

[0043] The dividing unit 102 divides the sensor data acquired by the data acquisition unit 101 into multiple sets according to the size of the values. Figure 3 etc. explain in detail how to set up and divide the set.

[0044] The relationship index calculation unit 103 calculates a relationship index representing the correlation of a group of sensors S, based on how the sensor data acquired by the sensors S of the group of two sensors S are distributed. Specifically, the relationship index calculation unit 103 calculates the relationship index based on the number of times the sensor data is divided into each of the multiple groups. This allows the relationship between the sensors to be quantified with simple processing. Figure 3 etc. will explain in detail the calculation method of the relationship index.

[0045] The map generation unit 104 uses the relationship index calculated by the relationship index calculation unit 103 for each group of sensors S to generate information representing the state of the mechanical equipment (in this embodiment, the waste incineration plant P) during a specified period, that is, a map. It will be explained in detail later that the map generated by the map generation unit 104 is an image in which a pattern corresponding to the value of the relationship index calculated for each group of sensors S is depicted on each segment corresponding to each group specified on the image plane. By generating such a map and displaying the output to the output unit 13, etc., the map generation unit 104 can enable the operator of the waste incineration plant P and the like to visually recognize the state of the waste incineration plant P. The above-mentioned map is information representing the state of the waste incineration plant P, and therefore the map generation unit 104 can also be called a state information generation unit.

[0046] (Partitioning of sensor data and calculation of relationship indicators)

[0047] based on Figure 3 The division of sensor data by the division unit 102 and the calculation of the relationship index by the relationship index calculation unit 103 will be described. Figure 3 This is a diagram showing an example of dividing sensor data and an example of calculating a relationship index.

[0048] exist Figure 3 121 shows the relationship between the value of the sensor data output by one of the sensors S1 to Sn, namely the sensor SA, and the value of the sensor data output by the other of the sensors S1 to Sn, namely the sensor SB. Figure 3 In 121, the sensor data values ​​output by sensor SA and the sensor data values ​​output by sensor SB are plotted on a coordinate plane as a pair. The left-right axis of this coordinate plane represents the magnitude of the output values ​​of sensor SA, while the up-down axis represents the magnitude of the output values ​​of sensor SB. For example, if the sensor data output by sensor SA during a predetermined period of operation of waste incineration plant P is a1 to a20, and the sensor data output by sensor SB during the same period is b1 to b20, the plotted points are (a1, b1) to (a20, b20).

[0049] In addition, Figure 3 In 121, the coordinate plane is divided into nine segments. These segments are set based on the size of the sensor data values ​​output by sensors SA and SB (three stages: Small, Middle, and Large). Specifically, nine segments are set, ranging from the segment where the sensor data values ​​output by sensors SA and SB are both "Small" (Small-Small) to the segment where the sensor data values ​​output by sensors SA and SB are both "Large" (Large-Large).

[0050] The dividing unit 102 may also divide the sensor data based on such a segment. In such a case, a threshold value for dividing the sensor data size is pre-set for each sensor S. For example, in the case of Figure 3 In the case of dividing the sensor data into three groups as in the example, the thresholds for dividing the sensor data into "Small" and "Middle" and the thresholds for dividing the sensor data into "Middle" and "Large" are set in advance. This allows the dividing unit 102 to divide the sensor data into "Small," "Middle," and "Large" groups based on their values. The method for setting the thresholds is not particularly limited; for example, the thresholds may be set based on information indicating the distribution of the sensor data (e.g., average, maximum, minimum, etc.).

[0051] The division unit 102 may also perform this division for each combination of sensors S1 to Sn. Furthermore, the sets into which the sensor data is divided may differ for each of sensors S1 to Sn. For example, the sensor data output by sensor S1 may be divided into two sets, while the sensor data output by sensor S2 may be divided into four or more sets.

[0052] In addition, the dividing unit 102 may also divide the sensor data by assigning the sensor data to a plurality of fuzzy sets. In this way, the sensor data near the boundary between a set and other sets can be appropriately assigned. For example, the dividing unit 102 may also use Figure 3 The membership function shown in 123 assigns the sensor data to three fuzzy sets: "Small," "Middle," and "Large." Membership functions can be created manually or automatically. In the case of automatic creation, the partitioning unit 102 may, for example, calculate the mean value and standard deviation σ of the sensor data and create the membership function based on the maximum, minimum, and average values ​​of the sensor data that fall within a range of ±3σ from the mean value.

[0053] The relationship index calculation unit 103 calculates the relationship index for each combination of sensors S1 to Sn based on the number of times the sensor data is divided into each set. Figure 3 In the example of 121, 9 of the 20 points shown are included in the "Middle"-"Middle" partition, 6 are included in the "Middle"-"Small" partition, and 2 are included in the "Small"-"Large" partition. Furthermore, one point is included in each of the "Small"-"Small," "Large"-"Small," and "Large"-"Middle" partitions, and no points are included in the other partitions.

[0054] In this case, the relationship index calculation unit 103 may also set the relationship index of each segment to be, for example, Figure 3 The value shown is 122. Figure 3 In the example of 122, the relationship index of the partition "Middle"-"Middle" containing the most drawn points is 0.405, and the relationship index of the partition "Middle"-"Small" containing the second most drawn points is 0.225. In addition, the relationship index of the partition "Small"-"Large" containing the third most drawn points is 0.1. For the partition containing only one point, since the relationship index is lower than the specified threshold, it is the same as the partition containing no points, and the relationship index is 0. Figure 3 In the example, the more times the sensor data is divided, the greater the value of the relationship index. Figure 4 The calculation method of such a relationship index will be described.

[0055] (Example of calculation method of relationship index)

[0056] Figure 4 : is a diagram showing an example of the division result of the values ​​of the sensor data outputted by the two sensors S. Specifically, Figure 4 In the example of 124, 30 of the sensor data output by sensor S1 are classified as "Small," 10 of the sensor data output by sensor S2 are classified as "Middle," and 20 of the sensor data output by sensor S2 are classified as "Large." Furthermore, the total number of sensor data output by sensor S2 is 120, of which 30 are classified as "Small," 40 are classified as "Middle," and 50 are classified as "Large."

[0057] On the other hand, Figure 4 In the example of 125, the division of the sensor data output by sensor S1 is the same as that of the example of 124, but the total number of sensor data output by sensor S2 is 450. Of the sensor data output by sensor S2, 150 are classified as "Small", 250 are classified as "Middle", and 50 are classified as "Large".

[0058] The relationship index calculation unit 103 may calculate the relationship index of each segment using the following mathematical formula (1).

[0059] (Relationship index) = (occupancy rate P) × (coverage rate C) × (ratio R) ... mathematical formula (1)

[0060] The occupancy rate P is the ratio of the sensor data belonging to "Small" for sensor S1 to the sensor data belonging to "Large" for sensor S2. Figure 4 In the example of 124, the occupancy rate P = 20 / 30.

[0061] The coverage ratio C is the ratio of the sensor data belonging to "Large" of sensor S2 to the sensor data belonging to "Small" of sensor S1. Figure 4 In the example of 124, the coverage ratio C = 20 / 50.

[0062] The ratio R is the ratio of the sensor data belonging to "Large" to the total number of sensor data of sensor S2. Figure 4 In the example of 124, the ratio R = 50 / 120.

[0063] According to the above situation, Figure 4 In the example of 124, the occupancy P = 20 / 30, the coverage C = 20 / 50, and the ratio R = 50 / 120, so the relationship index is 1 / 9. Figure 4 In the example of 125, the occupancy rate P = 20 / 30, the coverage rate C = 20 / 50, and the ratio R = 50 / 450, so the relationship index is 4 / 135.

[0064] In addition, multiplying the ratio R in the relationship index calculation formula is not essential, but by multiplying the ratio R, as in the example of 125, the relationship index value can be made appropriate even when the difference between the number of data in the second segment and the number of data in the other segments is large, so it is preferable. In addition, in the above mathematical formula (1), at least one of the occupancy rate P, the coverage rate C, and the ratio R can be multiplied by a weight. For example, by setting the weight of the occupancy rate P to a value greater than the weights of the coverage rate C and the ratio R, or by setting the weights of the coverage rate C and the ratio R to a value smaller than the weight of the occupancy rate P, it is possible to calculate a relationship index that prioritizes the occupancy rate P.

[0065] In addition, for example, in mathematical formula (1), the weight of the coverage C can also be set to zero to calculate the relationship index. In such a case, for the combination of opposite viewpoints (for example, the combination of sensors S2 and S1 relative to the combination of sensors S1 and S2), the weight of the occupancy rate P can also be set to zero to calculate the relationship index. Moreover, in such a case, the weights of other items in mathematical formula (1) can also be set to 1. In this way, for the combination of two viewpoints in the same sensor S (for example, if it is sensors S1 and S2, it is the combination of S1-S2 and S2-S1), different weights (or different calculation formulas) can also be used to calculate the relationship index. Even with such a structure, a correlation map representing the status of the waste incineration plant P can be generated.

[0066] (Generation of correlation mapping)

[0067] The map generation unit 104 generates a correlation map, which is information indicating the status of the waste incineration plant P, using the above-mentioned relationship index. Figure 5 The generation of the correlation map will be described. Figure 5 is a diagram showing an example of a correlation map representing the state of a waste incineration plant P. Figure 5 2 shows a normal version of the correlation map M114 and a simplified version of the correlation map M115.

[0068] The correlation map M114 is a map that represents the value of the relationship index of each divided combination using color for all combinations of sensors S1 to Sn. Figure 5 In , among all the combinations of sensors S1 to Sn, the corresponding parts are shown, which are the combination of sensors S14 to S19 and the combination of sensors S14 to S19 (it should be noted that the combination of the same sensor S is excluded). Figure 5 There are places where Small, Middle, and Large are abbreviated as S, M, and L respectively.

[0069] In the correlation map M114, a segment is defined for each sensor S combination, and each segment is further divided into nine subsegments. These nine subsegments correspond to the divisions of sensor data values ​​within the sensor S combination into Small, Medium, and Large. For example, the intersection of the column of sensor S14 and the row of sensor S15 constitutes the segment corresponding to the sensor S14-S15 combination. Furthermore, the upper left subsegment of the nine subsegments within this segment corresponds to the combination where both sensor data from sensor S14 and sensor S15 are "Small." The other subsegments similarly correspond to the sensor data divisions.

[0070] Figure 5The correlation map M114 is generated as follows: small segments where the relationship index is less than a preset threshold are not colored, the closer the relationship index is to 1, the closer it is to black, and the closer the relationship index is to 0, the closer it is to white. In other words, the correlation map M114 represents the relationship index using the pixel value in grayscale. In addition, the pattern used to represent the relationship index in the correlation map M114 is arbitrary and is not limited to Figure 5 For example, it can also be expressed by saturation, hue, brightness, or a combination thereof. If a rule for converting the relationship index into color is predetermined, the map generation unit 104 can determine the display color of each segment of the correlation map M114 according to the rule.

[0071] On the other hand, the correlation map M115 is an image obtained by depicting a pattern in a segment. The pattern for each segment is determined using a maximum of nine relationship indices corresponding to that segment. Specifically, the map generation unit 104 calculates the weighted sum of the relationship indices contained in a segment and sets the average value as the segment's relationship index value. The map generation unit 104 then determines the pattern corresponding to the segment's relationship index value (e.g., an image obtained by uniformly filling the entire segment with pixel values ​​corresponding to the relationship index value) as the pattern depicted in that segment.

[0072] The weights may be set to values ​​such that the distribution of sensor data for each combination of sensors S is reflected in the weighted sum. Figure 6 An example of weight setting is explained below. Figure 6 131 of the figure shows an example of a relationship index calculated for the combination of sensors S1 and S2, and 132 of the figure shows an example of setting a weight in the combination of sensors S1 and S2.

[0073] exist Figure 6 In the example of 132, the weight of the combination "Middle"-"Middle" is set to the largest value (specifically, 1.00). In addition, the weights of the combinations "Middle"-"Small", "Middle"-"Large", "Small"-"Middle", and "Large"-"Middle" are set to intermediate values ​​(specifically, 0.50). Furthermore, the weights of the other combinations are set to smaller values ​​(specifically, 0.25).

[0074] When these weights are applied, Figure 6The weighted sum average value of the relationship index of the combination of sensors S1 and S2 shown in 131 is (0.35×1.00+0.80×0.25) / 2=0.275. Similarly, for all combinations of sensors S, the weighted sum average value of the relationship index is calculated, and a pattern corresponding to the calculated relationship index value is drawn on the segment corresponding to each combination of sensors S. By this process, it is possible to draw Figure 5 The simplified version of the correlation map M115 is shown. The weights may be set to a common value for all combinations of sensors S, but are preferably adjusted to an optimal value for each combination of sensors S.

[0075] (Example of changes in correlation mapping)

[0076] Figure 7 is a diagram showing an example of a change in the correlation map generated as described above. Figure 7 In, with Figure 5 Likewise, a portion of the relevant map is selected and shown. Figure 7 The correlation map M116 shown is generated based on sensor data acquired during the stable period of the waste incineration plant P. According to the correlation map M116 , the sensor data of sensors S14 to S19 during the stable period are concentrated in “Middle”.

[0077] on the other hand, Figure 7 The correlation map M118 shown is generated based on sensor data acquired during an unstable period at the waste incineration plant P. Furthermore, the correlation map M117 is generated based on sensor data acquired during the period immediately preceding the unstable period. Correlation maps M116 to M118 show that the correlation between the magnitudes of the sensor data values ​​gradually decreases during the transition from a stable period to an unstable period.

[0078] More specifically, in correlation map M116, the "Middle"-"Middle" sub-segment is colored nearly black, while other sub-segments are uncolored. In correlation map M117, sub-segments other than "Middle"-"Middle" are colored, and some "Middle"-"Middle" sub-segments are uncolored. This trend is further developed in correlation map M118, with even fewer "Middle"-"Middle" sub-segments being colored.

[0079] As shown in the correlation maps M116 and M118 , it is possible to know from the correlation maps whether the waste incineration plant P is in a stable period or an unstable period. Furthermore, as shown in the correlation map M117 , it is also possible to know from the correlation map a precursor to an unstable period.

[0080] Therefore, the map generation unit 104 may also generate a reference map based on sensor data acquired during a stable period. This reference map serves as a reference for an operator, etc., using the operation support system 100, to determine the stable state of the waste incineration plant P. Furthermore, the map generation unit 104 may also generate a correlation map at any time based on sensor data acquired in real time during the operation of the waste incineration plant P, and have the output unit 13 display and output the generated correlation map along with the reference map. Thus, the operator, etc., can compare the reference map with the current correlation map to determine whether the waste incineration plant P is stable or is experiencing signs of instability.

[0081] Alternatively, the map generation unit 104 may generate a reference map based on sensor data acquired during an unstable period. In this case, the operator compares the reference map with the current correlation map and determines that there is a sign of instability if the current correlation map is close to the reference map.

[0082] (Processing Flow)

[0083] based on Figure 8 The flow of processing (information processing method) executed by the information processing device 1A will be described. Figure 8 1A is a flowchart showing an example of processing performed by the information processing device 1A. In addition, the following describes an example of generating a correlation map. In ST11, if sensor data measured in a waste incineration plant P that is operating normally is obtained, Figure 8 This is a flowchart for the process of creating a reference map. Furthermore, "normal operation" refers to a state where no abnormalities have occurred. For example, a state where the waste incineration plant P can continue to operate under automatic control without manual intervention, a state where the waste incineration plant P can continue to operate according to a predetermined operation plan, or a state where power generation is stable within a specified range is considered a "normal operation" state. Such a state can also be referred to as a stable state.

[0084] In ST11 (data acquisition step), the data acquisition unit 101 acquires sensor data from all sensors S. More specifically, the data acquisition unit 101 acquires a predetermined number of sensor data output during a predetermined period (eg, the last three minutes) via the input unit 12 .

[0085] In ST12, the segmentation unit 102 segments the sensor data acquired in ST11 according to the size of its values. For example, the segmentation unit 102 may segment the sensor data into three levels: Small, Middle, and Large, or two, or four or more. Furthermore, the segmentation unit 102 may segment the sensor data based on, for example, a threshold value or using a fuzzy set.

[0086] In ST13 (relationship index calculation step), relationship index calculation unit 103 calculates the relationship index for each segment based on the distribution of the sensor data acquired in ST11. As described above, relationship index calculation unit 103 calculates a relationship index corresponding to the number of data items included in each segment for each segment and each combination of segments. For example, relationship index calculation unit 103 may calculate the relationship index using the above-described mathematical formula (1).

[0087] In ST14 (state information generation step), the map generation unit 104 uses the relationship index calculated in ST13 to generate a correlation map representing the state of the waste incineration plant P. Specifically, the map generation unit 104 generates the correlation map by determining a plotting pattern corresponding to the relationship index for each combination of the Small, Middle, and Large segments in the segments of each combination of sensors S, thereby generating the correlation map.

[0088] In ST15, the map generator 104 causes the output unit 13 to display the correlation map generated in ST14. Furthermore, if the correlation map is already displayed on the output unit 13, the map generator 104 may update the displayed correlation map with the newly generated correlation map. This allows the displayed correlation map to always reflect the latest status of the waste incineration plant P.

[0089] In ST16, it is determined whether the data acquisition unit 101 has completed the processing. If it is determined that the data acquisition unit 101 has completed the processing (YES in ST16), Figure 8 The processing ends. On the other hand, if the data acquisition unit 101 determines that the processing is to continue (No in ST16), the process returns to ST11. The determination conditions of ST16 can be appropriately set in advance. For example, the data acquisition unit 101 may determine that the processing is to end when the waste incineration plant P is shut down. This allows operation support to be provided by displaying the relevant map until the waste incineration plant P is shut down.

[0090] (Implementation Method 2)

[0091] Another embodiment of the present invention will be described below. In addition, for the sake of convenience, components having the same functions as those described in the above embodiment are given the same reference numerals and their descriptions are not repeated. The same applies to the third embodiment and subsequent embodiments.

[0092] The information processing device 1B of this embodiment differs from the information processing device 1A of the above embodiment in that it automatically detects the occurrence of an abnormality or the presence of a sign thereof in the waste incineration plant P. Figure 9 The information processing device 1B will be described. Figure 91B and the flow of processing executed by the information processing device 1B. The information processing device 1B has the same configuration as the information processing device 1A (see FIG. Figure 1 ) are the same, so in Figure 9 2 shows the configuration of the control unit 10B.

[0093] like Figure 9 As shown, the information processing device 1B differs from the information processing device 1A in that it includes a stability index calculation unit 201 and an abnormality detection unit 202 instead of the map generation unit 104. In addition, the control unit 10B may also include the map generation unit 104.

[0094] The stability index calculation unit 201 calculates a stability index using the sensor data acquired by the data acquisition unit 101. The stability index is an indicator that indicates the degree of similarity between the distribution of sensor data when the waste incineration plant P is operating in a stable state and the distribution of sensor data acquired by the data acquisition unit 101. Since the stability index is information indicating the state of the waste incineration plant P, the stability index calculation unit 201 can also be referred to as a state information generation unit.

[0095] By including the stability index calculation unit 201, the information processing device 1B can numerically indicate whether the operating state of the waste incineration plant P is close to or far from a stable state (that is, whether it is unstable). The stability index can be displayed and output to the operator, etc., similar to the correlation map, and can also be used for abnormality detection, etc., as described below.

[0096] The stability index can be calculated, for example, as follows. Furthermore, before the stability index calculation unit 201 calculates the stability index, the relationship index calculation unit 103 calculates the relationship index using sensor data from when the waste incineration plant P is in a stable operating state. In the following description, the relationship index calculated using sensor data from when the waste incineration plant P is in a stable operating state is referred to as the stable-time relationship index. Furthermore, the aforementioned reference map can also be created using the stable-time relationship index.

[0097] First, the stability index calculation unit 201 calculates the fitness of the sensor data values ​​acquired by the data acquisition unit 101 with each segment based on the magnitude of the sensor data values. This sensor data represents the state of the waste incineration plant P, the target of the stability index calculation. Next, the stability index calculation unit 201 calculates the stability index for each segment based on the stability-time relationship index for each segment and the fitness calculated for that segment. The stability index calculation unit 201 then calculates the stability index for each combination of sensors S based on the stability index for each segment. The sum of the stability indices calculated for each combination is then used as the final stability index.

[0098] The following is based on Figure 10 The following describes the calculation of the stability relationship index of the combination of sensor S1 and sensor S2. Figure 6 An example of calculating a stability index in the case of the relationship indicator shown in 131. Figure 10 : is a diagram showing an example of calculating the stability index. Also, the values ​​of the sensor data of sensors S1 and S2 at time t after the calculation of the stability-related index are respectively represented by v1 and v2.

[0099] In such a case, if Figure 10 As shown in FIG141, the stability index calculation unit 201 calculates the degree of fitness of v1, which indicates the degree of fitness of each classification (Small, Middle, Large) of the value of sensor S1. In the calculation of fitness, the membership function of each classification of the value of sensor S1 can be used. Figure 10 In the example of 141, the fitness of v1 with the "Small", "Middle", and "Large" partitions is calculated to be 0.0, 0.7, and 0.3, respectively.

[0100] In addition, the stability index calculation unit 201 calculates the compatibility between the value v2 of the sensor data of the sensor S2 and each category (Small, Middle, Large) of the value in the sensor S2 in the same manner as described above. Figure 10 In the example of 141, the fitness of v2 with the "Small", "Middle", and "Large" partitions is calculated to be 0.4, 0.6, and 0.0, respectively.

[0101] Then, the stability index calculation unit 201 calculates the fitness for each division of the combination of large and small values ​​by multiplying the fitness for v1 by the fitness for v2. Figure 10 In the example of 141, the fitness of the "Middle"-"Small" partition is 0.7 × 0.4 = 0.28, and the fitness of the "Middle"-"Middle" partition is 0.7 × 0.6 = 0.42. Furthermore, the fitness of the "Large"-"Small" partition is 0.3 × 0.4 = 0.12, and the fitness of the "Large"-"Middle" partition is 0.3 × 0.6 = 0.18. The fitness of the other partitions is 0.0.

[0102] Next, the stability index calculation unit 201 calculates the stability index of each segment by multiplying the stability relationship index by the calculated fitness for that segment. Figure 6In the example of 131, the relationship index for the "Middle"-"Middle" partition is 0.35, the relationship index for the "Large"-"Small" partition is 0.8, and the relationship index for the other partitions is 0.0. Therefore, the stability index for the "Middle"-"Middle" partition is 0.35 × 0.42 = 0.147, and the stability index for the "Large"-"Small" partition is 0.12 × 0.8 = 0.096.

[0103] Then, the stability index calculation unit 201 continues to sum up the stability index of each division as the stability index when measuring the sensor data of v1 and v2. Figure 10 The stability index in the example of 142 is 0.147+0.096=0.243. The stability index calculation unit 201 performs the above processing on all combinations of sensors S and sets the sum of the stability indices calculated for each combination as the stability index at time t.

[0104] The stability index calculated in this manner increases in value as the distribution of sensor data acquired by the data acquisition unit 101 more closely resembles the distribution of sensor data when the waste incineration plant P is operating in a stable state. Therefore, the stability index can be said to indicate the degree of similarity between the distribution of sensor data when the waste incineration plant P is operating in a stable state and the distribution of sensor data acquired by the data acquisition unit 101.

[0105] In addition, the stability index calculation unit 201 can also calculate the stability index by considering the temporal changes of the sensor data. For example, the stability index calculation unit 201 can also calculate the centroid position of the temporal sensor data and use the centroid position to calculate the stability index. Figure 10 143 for explanation. Figure 10 In the example of 143 , it is assumed that sensor data ( v11 , v21 ), ( v12 , v22 ), and ( v13 , v23 ) are measured by sensors S1 and S2 at time t1 , t2 , and t3 .

[0106] In this case, the stability index calculation unit 201 calculates the center of gravity (v1', v2') of (v11, v21), (v12, v22), and (v13, v23). The stability index calculation unit 201 can then calculate the stability index for times t1 to t3 using the same calculations as in the above example of calculating the stability index using v1 and v2.

[0107] In this way, the stability index calculation unit 201 can also (1) use the plurality of sensor data measured during a predetermined period to calculate data indicating the state of the waste incineration plant P during that period (in the above example, the coordinates of the center of gravity), and (2) use this data to calculate the stability index. Thus, the stability index calculation unit 201 can calculate the stability index by taking into account the temporal changes in the sensor data during that period.

[0108] Alternatively, instead of the stability index, an instability index may be calculated to indicate the overall degree of similarity between the distribution of sensor data acquired by the data acquisition unit 101 and the distribution of sensor data when the waste incineration plant P is in an unstable operating state. Alternatively, information indicating the degree of divergence between the distribution of sensor data acquired by the data acquisition unit 101 and the distribution of sensor data when the waste incineration plant P is in an unstable operating state may be used as the stability index. In this manner, the operating state used as the reference for calculating the stability index or the instability index can be either a stable state or an unstable state.

[0109] The abnormality detection unit 202 detects abnormalities in the waste incineration plant P based on the stability index calculated by the stability index calculation unit 201. Furthermore, the abnormalities that the abnormality detection unit 202 detects include not only unstable states of the waste incineration plant P but also states that indicate a tendency for the state of the waste incineration plant P to become unstable.

[0110] The abnormality detection unit 202 may also determine that an abnormality has occurred when the stability index is below a predetermined threshold. Furthermore, the abnormality detection unit 202 may determine the presence of an abnormality based on the rate of change of the stability index. For example, the abnormality detection unit 202 may determine that an abnormality has occurred when the stability index decreases sharply over a short period of time, or when the stability index decreases continuously over a predetermined period of time.

[0111] Furthermore, if the abnormality detection unit 202 determines that an abnormality may have occurred, it notifies the operator or the like of this fact. The method of notification is not particularly limited. For example, the abnormality detection unit 202 may make the notification by causing the output unit 13 to display information indicating that an abnormality has been detected. Alternatively, another processing module may be used to perform abnormality determination and notification.

[0112] Next, the flow of the processing (information processing method) performed by the information processing device 1B will be described. Figure 8 The processes of ST11 and ST16 are the same, so the processes of ST22 to ST25 will be described below.

[0113] In ST22, the stability index calculation unit 201 calculates the suitability of each segment with respect to the sensor data acquired in ST21. Specifically, based on Figure 10 As described above, the stability index calculation unit 201 calculates the compatibility between the sensor data acquired in ST21 and the three categories of "Small," "Middle," and "Large" for each combination of sensors S. The stability index calculation unit 201 then multiplies the calculated compatibility and calculates the compatibility between the sensor data acquired in ST21 and the nine categories of "Small"-"Small" to "Large"-"Large."

[0114] In ST23 (state information generation step), the stability index calculation unit 201 uses the fitness calculated in ST22 to calculate a stability index representing the state of the waste incineration plant P. Specifically, based on Figure 10 As described above, the stability index calculation unit 201 calculates the stability index by multiplying the fitness calculated in ST22 by the stability-related index.

[0115] Then, in ST24, the abnormality detection unit 202 determines whether an abnormality exists based on the stability index calculated in ST23. If it is determined in ST24 that an abnormality exists (YES in ST24), the process proceeds to ST25, where the operator is notified that the abnormality detection unit 202 has detected an abnormality. On the other hand, if it is determined in ST24 that there is no abnormality (NO in ST24), the process proceeds to ST26.

[0116] Furthermore, in ST23, in addition to the overall stability index, the stability index calculation unit 201 may also calculate the stability index for a combination of some sensors S. In this case, in ST24, if the overall stability index determines that there is no abnormality, the abnormality detection unit 202 may re-determine the presence of an abnormality based on the stability index for the combination of some sensors S. This allows abnormality detection even when signs of an abnormality are difficult to detect based on the overall stability index but the output values ​​of some sensors S show signs of an abnormality.

[0117] (About machine learning)

[0118] The anomaly detection unit 202 may also detect anomalies in the waste incineration plant P using a learned model constructed by machine learning the relationship between the relationship index calculated by the relationship index calculation unit 103 and the status of the waste incineration plant P. In this case, the stability index calculation unit 201 is omitted, and the anomaly detection unit 202, which inputs the relationship index into the learned model and outputs information indicating whether the waste incineration plant P is abnormal, functions as a status information generation unit.

[0119] In the machine learning described above, for example, the teacher data may include relationship indices calculated from sensor data acquired during normal operation of the waste incineration plant P and relationship indices calculated from sensor data acquired a predetermined time before an abnormality occurred in the waste incineration plant P. By inputting the relationship indices calculated by the relationship indices calculation unit 103 into a learned model constructed through such machine learning, it is possible to obtain output values ​​indicating the likelihood that the waste incineration plant P is operating normally and output values ​​indicating the likelihood that an abnormality will occur after the predetermined time.

[0120] In such a case, the abnormality detection unit 202 can detect an abnormal operating state, for example, when the output value indicating the possibility of normal operation is less than a predetermined threshold. Alternatively, the abnormality detection unit 202 can detect the presence of a sign of abnormality, for example, when the output value indicating the possibility of abnormality occurring after a predetermined time is greater than a threshold.

[0121] (Implementation 3)

[0122] In this embodiment, an example is described in which the value of subsequent sensor data is predicted based on the transition of the value of acquired sensor data, and abnormality detection is performed based on the predicted value of the sensor data. Figure 11 An information processing device 1C according to this embodiment will be described. Figure 11 1C and a diagram showing an overview of a method for predicting sensor data values. The configuration of the information processing device 1C other than the control unit 10C is similar to that of the information processing device 1A (see FIG. Figure 1 ) are the same, so in Figure 11 2 shows the configuration of the control unit 10C.

[0123] The information processing device 1C is similar to the information processing device 1B of the second embodiment (see Figure 9 ) differs from the previous one in that it includes a data prediction unit 301. The data prediction unit 301 calculates predicted values ​​for future sensor data based on the changes in the sensor data values ​​acquired by the data acquisition unit 101. The partitioning unit 102 then partitions these predicted values, and the relationship index calculation unit 103 calculates relationship indices based on these partitions. This allows the stability index calculation unit 201 to calculate a stability index representing the future state of the waste incineration plant P based on these predicted values. Using this stability index, measures can be taken to prevent instability in the operating state of the waste incineration plant P before it becomes unstable.

[0124] exist Figure 11The S1-S4 segment of the correlation map M311 depicts the coordinates of the combined sensor data values ​​output by sensor S1 and sensor S4. Furthermore, as indicated by the arrows in the figure, the closer the plotted point corresponds to data later in the time series, the closer it is to the lower right of the segment. Therefore, based on this trend, the data prediction unit 301 can predict the position of a future point (a combination of sensor data values ​​output by sensors S1 and S4). For example, the data prediction unit 301 can also calculate the coordinates of future plotted points based on an approximate curve derived from the coordinates of the plotted points in the time series up to the current moment, and use this approximate curve to calculate the coordinates of the future plotted points.

[0125] Alternatively, for example, the data prediction unit 301 may calculate predicted values ​​for future sensor data based on time-series sensor data up to the current time (trend data indicating a tendency of time-series changes in sensor data). In this case, for example, sensor data measured within the most recent specified time (e.g., 30 minutes) may be used.

[0126] The processing after the predicted value calculation is the same as that in Embodiment 2. Furthermore, the control unit 10C of the information processing device 1C may include the map generation unit 104 described in Embodiment 1. In this case, a correlation map representing the future state of the waste incineration plant P can be created based on the predicted value calculated by the data prediction unit 301.

[0127] (Implementation 4)

[0128] In this embodiment, an information processing device 1D having a function of detecting a failure of a sensor S and a function of determining the cause of an abnormality in a waste incineration plant P when the abnormality is detected will be described. Figure 12 1D. The information processing device 1D has the same configuration as the information processing device 1A (see FIG. 1A ) except for the control unit 10D. Figure 1 ) are the same, so in Figure 12 2 shows the configuration of the control unit 10D.

[0129] The information processing device 1D is similar to the information processing device 1B of the second embodiment (see Figure 9 ), differs from the information processing device 1A in that it includes a sensor failure detection unit 401, a cause determination unit 402, and a plant control unit 403. Furthermore, since sensor failure detection unit 401, cause determination unit 402, and plant control unit 403 do not work together, any one of them may be omitted. Alternatively, cause determination unit 402 may be provided while plant control unit 403 is omitted. Furthermore, sensor failure detection unit 401 may also be included in the information processing device 1A of Embodiment 1.

[0130] The sensor failure detection unit 401 detects a faulty sensor S based on the relationship indices calculated for each group of multiple sensors S. If a faulty sensor S is included in the multiple sensors S, this detection can be performed by excluding the sensor data output by the faulty sensor S from the normal range using the relationship indices calculated. For example, the sensor failure detection unit 401 may have the stability index calculation unit 201 calculate a stability index for each sensor S and detect sensors S whose calculated stability index falls below a threshold as faulty sensors S.

[0131] The cause determination unit 402 determines the cause of the waste incineration plant P falling into the state indicated by the stability index calculated by the stability index calculation unit 201 based on the relationship indicators calculated for each group of the plurality of sensors S. This allows appropriate measures to be taken according to the cause, improving the state of the waste incineration plant P.

[0132] More specifically, when the abnormality detection unit 202 determines that an abnormality exists, the cause determination unit 402 determines the cause of instability. The plant control unit 403 then performs control corresponding to the determination result of the cause determination unit 402, thereby stabilizing the state of the waste incineration plant P. Alternatively, the cause determination unit 402 may display and output the determination result to notify an operator, etc. In this case, the state of the waste incineration plant P is stabilized through manual operation by the operator.

[0133] The correlation between the relationship indicator and the cause is used to determine the cause. For example, if the state of the waste incineration plant P becomes unstable, if the relationship between the post-combustion grate temperature and the primary combustion air flow rate, as indicated by the relationship indicator, is "Small" and "Large," then it can be considered that the cause is insufficient waste supply to the incinerator. Therefore, if the sensor data output by sensor S representing the post-combustion grate temperature and the sensor data output by sensor S representing the primary combustion air flow rate are in a relationship of "Small" and "Large," the cause determination unit 402 can determine that the cause is insufficient waste supply to the incinerator. In this way, the relationship indicator can be used as information representing the overall state of the waste incineration plant P, and the analysis results of each component of the relationship indicator can also be used to determine the cause of the instability, etc.

[0134] The plant control unit 403 controls the operation of the waste incineration plant P. For example, the plant control unit 403 may control the operation of the waste incineration plant P by controlling a control device that controls the operation of various devices included in the waste incineration plant P.

[0135] Furthermore, when the relationship indicators calculated for each group of multiple sensors S meet predetermined conditions, the plant control unit 403 functions as a device control unit that causes the waste incineration plant P to execute an action to return the waste incineration plant P to a normal state. This automatically stabilizes the state of the waste incineration plant P.

[0136] Specifically, as described above, the plant control unit 403 stabilizes the state of the waste incineration plant P by performing control corresponding to the determination result of the cause determination unit 402. Since the cause determination unit 402 determines the cause corresponding to the condition when each relationship indicator calculated for each group of multiple sensors S meets the specified condition, the plant control unit 403 performs control when the above conditions are met. It is sufficient to predetermine the control content for each cause. For example, when the cause is insufficient supply of garbage to the incinerator, the plant control unit 403 only needs to control the supply of garbage to the incinerator. In addition, the cause determination unit 402 can also be omitted, and the plant control unit 403 can determine whether the specified conditions are met.

[0137] (Variation)

[0138] The execution subject of each process described in the above embodiments can be changed as appropriate. For example, one or more devices other than the information processing device 1A may be made to execute the process. Figure 8 For example, another information processing device may execute the processes of ST11 and ST12, and another information processing device may execute the process of ST13. In this case, the information processing device 1A acquires the relationship index calculated by the other information processing device to generate a map. Figure 9 The same is true for the flowcharts, and the execution subject of each process can be changed appropriately.

[0139] Furthermore, the structure of dividing sensor data into sets such as Large, Middle, and Small is one method for expressing the distribution of sensor data, and can be replaced with another method that can express the distribution of sensor data.

[0140] (Software-based implementation example)

[0141] The control modules of the information processing devices 1A to 1D (particularly, the units included in the control units 10A to 10D) may be implemented by logic circuits (hardware) formed on an integrated circuit (IC chip) or the like, or by software.

[0142] In the latter case, the information processing devices 1A to 1D include a computer that executes commands of software, i.e., a program (information processing program), that implements each function. The computer includes, for example, one or more processors and a computer-readable storage medium that stores the program. Furthermore, in the computer, the processor reads the program from the storage medium and executes it, thereby achieving the purpose of the present invention. As the processor, for example, a CPU (Central Processing Unit) can be used. As the storage medium, a "non-transitory tangible medium" can be used, such as a magnetic tape, an optical disk, a card, a semiconductor memory, a programmable logic circuit, etc., in addition to ROM (Read Only Memory). Furthermore, a RAM (Random Access Memory) or the like for developing the program may be included. Furthermore, the program may be provided to the computer via any transmission medium capable of transmitting the program (such as a communication network or broadcast waves). Furthermore, one embodiment of the present invention may be implemented as a data signal embedded in a carrier wave, embodied by electronically transmitting the program.

[0143] The present invention is not limited to the above-described embodiments, and various modifications can be made within the scope of the claims. Embodiments obtained by appropriately combining technical solutions disclosed in different embodiments are also included in the technical scope of the present invention.

[0144] Description of Reference Numerals

[0145] 1A~1D-information processing device; 101-data acquisition unit; 102-division unit; 103-relationship index calculation unit; 104-mapping generation unit (status information generation unit); 201-stability index calculation unit (status information generation unit); 301-data prediction unit; 401-sensor fault detection unit; 402-cause determination unit; 403-plant control unit (equipment control unit); P-waste incineration plant (mechanical equipment); S-sensor.

Claims

1. An information processing device, characterized in that have: a data acquisition unit that acquires sensor data, the sensor data being output values ​​of a plurality of sensors provided in the machine during a predetermined period, or numerical values ​​calculated using the output values; a relationship index calculation unit that calculates a relationship index indicating correlation of a group of sensors consisting of two of the plurality of sensors based on a distribution status of each sensor data acquired from sensors of the group; a state information generating unit for generating information indicating a state of the machine during the predetermined period using the relationship index calculated for each of the plurality of sensor groups; as well as A division unit, which divides the sensor data into multiple sets according to the size of the sensor data. The relationship index calculation unit calculates the relationship index based on the number of times the sensor data is divided into each of the plurality of sets. The state information generating unit generates an image as information representing the state of the mechanical device, in which patterns corresponding to the values ​​of the relationship indicators calculated for each group of the plurality of sensors are drawn in each segment corresponding to each group specified on an image plane.

2. The information processing device according to claim 1, wherein The division unit divides the sensor data by assigning the sensor data to a plurality of fuzzy sets.

3. An information processing device, characterized in that have: a data acquisition unit that acquires sensor data, the sensor data being output values ​​of a plurality of sensors provided in the machine during a predetermined period, or numerical values ​​calculated using the output values; a relationship index calculation unit that calculates a relationship index indicating correlation of a group of sensors consisting of two of the plurality of sensors based on a distribution status of each sensor data acquired from sensors of the group; a state information generating unit for generating information indicating a state of the machine during the predetermined period using the relationship index calculated for each of the plurality of sensor groups; as well as A division unit, which divides the sensor data into multiple sets according to the size of the sensor data. The relationship index calculation unit calculates the relationship index based on the number of times the sensor data is divided into each of the plurality of sets. The state information generating unit generates, as information indicating the state of the machine, information indicating a similarity between a distribution state of the sensor data acquired by the data acquiring unit and a distribution state of the sensor data when the machine is in a reference operating state.

4. The information processing device according to any one of claims 1 to 3, characterized in that The information processing device includes a data prediction unit that calculates a predicted value of subsequent sensor data based on a change in the value of the sensor data acquired by the data acquisition unit. The state information generating unit generates information indicating a future state of the machine based on the predicted value.

5. The information processing device according to any one of claims 1 to 3, characterized in that A cause determination unit is provided for determining a cause of the machine being in the state indicated by the information generated by the state information generation unit based on the relationship indicators calculated for each of the plurality of sensor groups.

6. The information processing device according to any one of claims 1 to 3, characterized in that A device control unit is provided for causing the mechanical device to execute an operation for transitioning the mechanical device to a normal state when each relationship index calculated for each group of the plurality of sensors satisfies a predetermined condition.

7. The information processing device according to any one of claims 1 to 3, characterized in that: A sensor failure detection unit is provided for detecting a sensor in which a failure has occurred based on each relationship index calculated for each group of the plurality of sensors.

8. An operation assisting system that assists the operation of mechanical equipment, characterized in that: The operation assistance system includes: information processing device; and Multiple sensors are provided on the mechanical equipment. The information processing device acquires sensor data, wherein the sensor data is output values ​​of the plurality of sensors during a prescribed period, or numerical values ​​calculated using the output values. The information processing device divides the sensor data into a plurality of sets according to the magnitude of the sensor data, calculates a relationship index representing the correlation of a sensor group consisting of two of the plurality of sensors based on the distribution of the sensor data acquired from the sensors, and generates and outputs information representing the state of the mechanical equipment during the prescribed period using the relationship index calculated for each of the plurality of sensor groups. The information processing device calculates the relationship index based on the number of times the sensor data is divided into each of the plurality of sets. An image is generated as information representing the state of the mechanical device, in which patterns corresponding to the values ​​of the relationship indicators calculated for each group of the plurality of sensors are drawn on each segment corresponding to each group specified on an image plane.

9. An information processing method, which is executed by an information processing device, characterized in that: The information processing method comprises: a data acquisition step of acquiring sensor data, wherein the sensor data is output values ​​of a plurality of sensors provided in the mechanical equipment during a predetermined period, or numerical values ​​calculated using the output values; The division step divides the above sensor data into multiple sets according to the size of its value. a relationship index calculating step of calculating a relationship index indicating correlation of a group of sensors consisting of two of the plurality of sensors based on a distribution status of each sensor data acquired from sensors of the group; and a state information generating step of generating information indicating the state of the mechanical device during the predetermined period using the relationship index calculated for each of the plurality of sensor groups; In the relationship index calculation step, the relationship index is calculated based on the number of times the sensor data is divided into each of the plurality of sets. In the above-mentioned status information generation, the following image is generated as information representing the status of the above-mentioned mechanical equipment, in which patterns corresponding to the values ​​of the above-mentioned relationship indicators calculated for each group of the plurality of the above-mentioned sensors are drawn onto each segment corresponding to the above-mentioned groups specified on the image plane.

10. A computer-readable storage medium storing an information processing program for causing a computer to function as the information processing device according to claim 1 or 3. The information processing program is for causing a computer to function as the data acquisition unit, the relationship index calculation unit, the status information generation unit, and the division unit.

11. An operation assisting system for assisting the operation of mechanical equipment, characterized in that: The operation assistance system includes: information processing device; and Multiple sensors are provided on the mechanical equipment. The information processing device acquires sensor data, wherein the sensor data is output values ​​of the plurality of sensors during a prescribed period, or numerical values ​​calculated using the output values. The information processing device divides the sensor data into a plurality of sets according to the magnitude of the sensor data, calculates a relationship index representing the correlation of a sensor group consisting of two of the plurality of sensors based on the distribution of the sensor data acquired from the sensors, and generates and outputs information representing the state of the mechanical equipment during the prescribed period using the relationship index calculated for each of the plurality of sensor groups. The information processing device calculates the relationship index based on the number of times the sensor data is divided into each of the plurality of sets. Information indicating the state of the machine is generated, the information indicating the similarity between the distribution of the acquired sensor data and the distribution of the sensor data when the machine is in a reference operating state.

12. An information processing method, which is executed by an information processing device, characterized in that: The information processing method comprises: a data acquisition step of acquiring sensor data, wherein the sensor data is output values ​​of a plurality of sensors provided in the mechanical equipment during a predetermined period, or numerical values ​​calculated using the output values; The division step divides the above sensor data into multiple sets according to the size of its value. a relationship index calculating step of calculating a relationship index indicating correlation of a group of sensors consisting of two of the plurality of sensors based on distribution conditions of respective sensor data acquired from sensors of the group; and a state information generating step of generating information indicating the state of the mechanical device during the predetermined period using the relationship index calculated for each of the plurality of sensor groups; In the relationship index calculation step, the relationship index is calculated based on the number of times the sensor data is divided into each of the plurality of sets. In the state information generation, information indicating the state of the mechanical device is generated, the information indicating the similarity between the distribution of the sensor data acquired in the data acquisition step and the distribution of the sensor data when the mechanical device is in a reference operating state.

Images