Method for displaying system balance fluctuations and their signs, and information processing device and program using said method

Abstract

To provide a system balance fluctuation indication method capable of accurately indicating a system balance fluctuation and its sign, and an information processor and program using the system balance fluctuation indication method.SOLUTION: An information processor 1 is employed in a method of indicating a system balance fluctuation and its sign. The information processor 1 includes control means 11 that uses time-series data items of N items obtained from a system to calculate degrees of contribution, divides the calculated degree of contribution on each of the items relative to each date by a sum total of the degrees of contribution on all the items in order to calculate a ratio of the degree of contribution on each item, assigns a color to each of the items for which the ratio has been calculated, layers assigned colors on the basis of the calculated degrees of contribution on the respective items relative to each date, and thus indicates the degrees of contributions as a percentage on the respective items relative to each date in colors.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] This invention relates to a system balance fluctuation display method, which uses the multivariate analysis technique "MT method" to distinguish between normal and abnormal states of an object using the "Mahalanobis distance," a distance that takes into account the correlation between multiple variables, as well as an information processing device and program that use this method. [Background technology]

[0002] The "MT method" is a multivariate analysis method that uses the "Mahalanobis distance," a distance that takes into account the correlation between multiple variables, to distinguish between the normal and abnormal states of a target. First, a correlation coefficient matrix relating to the correlation coefficients between each item of the reference space data is created, and then the Mahalanobis distance D M This allows us to understand the state change of the target space relative to the reference space.

[0003] In addition, the Mahalanobis distance D M For the calculation method of , see, for example, the technical literature "Try and Learn! Introduction to MT System Analysis Methods for Quality Engineering" (Nikkan Kogyo Shimbun, first published May 30, 2012), and the Mahalanobis distance D using complex number data. M The calculation method of is known based on, for example, the publication of the patent document "Method for analyzing and diagnosing the state of a system having multiple variables, and information processing device using said method" (Patent Application No. 2020-045873), and therefore will not be described here.

[0004] In recent years, this Mahalanobis distance D M Many methods and devices have been developed to determine whether or not an abnormality exists in a monitored object using the above method.

[0005] For example, in Patent Document 1, contact between a substrate transfer machine that transfers substrates from a pod, which is a storage container that stores substrates (wafers), to a boat, which is a substrate holder, or a substrate on the substrate transfer machine, and a substrate holder or a substrate on the substrate holder is detected by the Mahalanobis distance D MThis publication discloses a substrate processing apparatus that can detect minor abnormalities before a failure occurs by detecting the abnormalities with high accuracy using the above method.

[0006] In addition, in Patent Document 2, the Mahalanobis distance D M Calculate the Mahalanobis distance D M The present invention discloses a configuration of a production equipment abnormality diagnosis system that can easily diagnose production equipment by comparing the value of a production equipment signal with a threshold value to determine whether or not there is an abnormality in the production equipment, even when there are many items that indicate the state of the production equipment.

[0007] In addition, in Patent Document 3, different unit spaces are created based on the state quantities acquired during the start-up operation period and the rated speed operation period, which are different operating states, and the Mahalanobis distance D M When calculating the Mahalanobis distance D M When determining whether the plant is normal or not based on the above, one of two unit spaces is selected depending on whether the evaluation period is the start-up operation period or the rated speed operation period, and the Mahalanobis distance D M The present invention discloses a configuration for monitoring the state of a plant by determining whether the state is normal or abnormal by calculating the above. [Prior art documents] [Patent documents]

[0008] [Patent Document 1] Re-tabled publication 2020 / 157967

[0009] [Patent Document 2] JP 2018-92437 A

[0010] [Patent Document 3] Patent No. 5031088

[0011] All of Patent Documents 1 to 3 show "time axis innovations" for appropriately setting the reference space, and realize normal / abnormal diagnosis of equipment and systems.

[0012] On the other hand, in recent years, in analytical methods for determining whether something is normal or abnormal, the process and basis for the judgment are often kept as a black box, and there is a demand for transparency in the process and basis for the judgment. This is also called explainable AI (XAI). In the MT method, the Mahalanobis distance D M When an abnormality or accident is detected, it is possible to determine which item has the strongest influence and which item is the closest to the Mahalanobis distance D M There is a concept called "degree of contribution" that indicates how much a problem contributes to an increase in quality, and it is expected that this "degree of contribution" will be used to make the process and basis for determining whether something is normal or abnormal more transparent. This "degree of contribution" is calculated using the following procedure, as described in the aforementioned technical document, "Try and Learn! Introduction to Quality Engineering MT System Analysis Methods."

[0013] The Mahalanobis distance D M Next, remove one item and calculate the Mahalanobis distance D M Then, the Mahalanobis distance D calculated for all N items is calculated. M " and "The Mahalanobis distance D for N-1 items, excluding one item M This difference is the Mahalanobis distance D M By performing these steps for each item, the Mahalanobis distance D M The "contribution" to is calculated. Hereafter, when we express "contribution", we mean "Mahalanobis distance D M This indicates the degree of contribution to the

[0014] Conventionally, users such as experts have used this "degree of contribution" to analyze accidents and abnormalities and try to understand the factors that caused the accidents and abnormalities, as described below.

[0015] As shown in Figure 7, a contribution stack diagram is created that displays the contributions of each item for each date and time. The vertical axis of this contribution stack diagram represents the contribution value, and the horizontal axis represents the date and time. Dates and times when abnormalities or accidents occur have higher contribution stacks than other dates and times, making it possible to identify the dates and times when abnormalities or accidents occur. When the contribution stack diagram is displayed on the display of an information processing device such as a PC, a pie chart is displayed when the cursor is placed over the stack for a date and time of interest (in the figure, May 1st, 18:00). From this pie chart, users such as experts can understand the contribution ratio (breakdown) for that date and time. Therefore, users can understand the changes in the contribution ratio for each date and time by moving the cursor and looking continuously. Summary of the Invention [Problem to be solved by the invention]

[0016] However, in order to obtain the contribution stack needed to perform the above-mentioned factor analysis, it was necessary to select the appropriate items, and even highly skilled and experienced users had to go through a process of trial and error, redoing their item selection until they determined the correct answer based on their own experience. This process required comparing multiple MT analysis results, and it was difficult to remember the time-series changes in all of the above-mentioned pie charts, which led to a prolonged process and inevitable abandonment.

[0017] Therefore, the results of the contribution analysis are calculated using the Mahalanobis distance D M The only thing that can be used is to use it in areas with high values ​​of the Mahalanobis distance D M For periods when the value is small and cannot be confirmed unless it expands, the idea of ​​using the change in contribution to detect signs of accidents or abnormalities, or fluctuations in the system balance, was not often conceived.

[0018] In order to address the above-mentioned problems, the present invention provides a method for displaying system balance fluctuations and their signs, which is capable of accurately displaying system balance fluctuations and their signs, as well as an information processing device and program using said method. [Means for solving the problem]

[0019] In order to achieve the above object, the invention according to claim 1 comprises: A method for displaying fluctuations in system balance and their signs, Calculating the contribution using N items of time series data obtained from the system; A step of calculating a ratio of the contribution degree of each item by dividing the calculated contribution degree of each item by the total value of the contribution degrees of all items for each date and time; assigning a color to each of the percentage-calculated items; The method includes a step of displaying the contribution of each item for each date and time as a percentage by piling up the assigned color based on the calculated contribution rate.

[0020] The invention according to claim 2 is as follows: In a method for displaying fluctuations in system balance and their signs, Calculating the contribution using N items of time series data obtained from the system; The method includes a step of arranging each item on a vertical axis, color-coding the time series change in the magnitude of the contribution of each item according to the magnitude of the contribution, and arranging and displaying the change on a horizontal axis.

[0021] The invention according to claim 3 is as follows: A step of calculating a ratio of the contribution degree of each item for each date and time by dividing the calculated contribution degree of each item by the total value of the contribution degrees of all items; The method described in claim 2 includes a step of arranging each item on the vertical axis, color-coding the time series changes in the calculated ``contribution rate of each item'' according to the size of the contribution rate, and displaying them on the horizontal axis.

[0022] The invention according to claim 4 is as follows: An information processing device used in a method for displaying fluctuations in system balance and signs thereof, The information processing device includes: Calculate the contribution using N items of time series data obtained from the system, For each date and time, the calculated contribution rate for each item is divided by the total value of the contribution rates for all items to calculate the contribution rate for each item; Assign a color to each item whose percentage was calculated, The information processing device has a control means for displaying the contribution of each item for each date and time as 100% by piling up the assigned color based on the calculated contribution rate.

[0023] The invention according to claim 5 is as follows: A program for operating an information processing device having a control means, which is used in a method for extracting fluctuations in system balance and signs thereof, The program causes the control means of the information processing device to: Calculate the contribution using N items of time series data obtained from the system, For each date and time, the calculated contribution degree for each item is divided by the total value of the contribution degrees for all items to calculate the proportion of the contribution degree for each item; Assign a color to each item whose percentage was calculated, The program displays the contribution of each item for each date and time as a percentage by piling up the assigned color based on the calculated contribution rate.

[0024] The invention according to claim 6 is as follows: An information processing device used in a method for extracting fluctuations in system balance and signs thereof, The information processing device includes: Calculate the contribution using N items of time series data obtained from the system, The information processing device has a control means for arranging each item on the vertical axis, color-coding the time series change in the magnitude of the contribution of each item according to the magnitude of the contribution, and displaying them on the horizontal axis.

[0025] The invention according to claim 7 is as follows: A program for operating an information processing device having a control means, which is used in a method for extracting fluctuations in system balance and signs thereof, The program causes the control means of the information processing device to: Calculate the contribution using N items of time series data obtained from the system, The program was designed to arrange each item on the vertical axis, and display the time series changes in the degree of contribution of each item, color-coded according to the degree of contribution, on the horizontal axis.

[0026] The invention according to claim 8 is as follows: The control means divides the calculated contribution degree for each item by the total value of the contribution degrees for all items for each date and time to calculate a ratio of the contribution degree for each item for each date and time; The information processing device according to claim 6 arranges each item on the vertical axis, and the time series changes in the calculated "proportion of contribution of each item" are color-coded according to the magnitude of the proportion of contribution and displayed on the horizontal axis.

[0027] The invention according to claim 9 is as follows: The program causes the control means of the information processing device to For each date and time, the calculated contribution degree for each item is divided by the total value of the contribution degrees for all items to calculate the proportion of the contribution degree for each item for each date and time; The program described in claim 7 arranges each item on the vertical axis, and the time series changes in the calculated "proportion of contribution of each item" are color-coded according to the magnitude of the proportion of contribution and displayed on the horizontal axis. [Effects of the Invention]

[0028] In this invention, by displaying the contribution stack as 100%, it is possible to check the ratio and trend of the contribution for each date and time (=the time-series change in the contribution ratio can be seen), which was previously impossible to check unless the contribution stack was enlarged because the values ​​were small. At the same time, it is also possible to check the change points in the contribution ratio of each item, which makes it possible to grasp fluctuations in the system balance and their signs.

[0029] This invention extracts predictive items and displays them as a predictive item extraction diagram. This fusion of information related to each item allows for visualization of insights and anomalies (contradictions). As described above, an "information catalyst" is defined as a space where information is input and something more is created. The predictive item extraction diagram functions as an information catalyst. Furthermore, by displaying each item grouped by function, the information catalyst effect is enhanced, inducing visual insights (contradictions) when assembling abnormality factors and their mechanisms and logic. It also promotes a bird's-eye view of the system, as if viewing a real system. Furthermore, this bird's-eye view leads to the discovery of not only individual issues but also issues and contradictions that span systems. It also leads to the detection of unexpected abnormalities and the discovery of new phenomena. [Brief explanation of the drawings]

[0030] [Figure 1] FIG. 10 is a contribution stack diagram showing, in chronological order, the contribution of each item to the Mahalanobis distance at each date and time in the first embodiment of the present invention. [Figure 2] This is a 100% display diagram of a contribution stack in which the contribution of each item to the Mahalanobis distance at each date and time is stacked and the stacked contribution is displayed as a 100% ratio in the first embodiment of the present invention. [Figure 3] FIG. 10 is a predictor item extraction diagram created from a contribution stack diagram showing the contribution of each item to the Mahalanobis distance at each date and time in the first embodiment of the present invention in a time series. [Figure 4] FIG. 10 is a predictor item extraction diagram created from a contribution stack 100% display diagram showing the contribution of each item to the Mahalanobis distance at each date and time in the first embodiment of the present invention. [Figure 5] This is a comparison diagram in which Figure 1 and Figure 3, which was created from Figure 1, are arranged vertically with the time axes aligned and placed on the left, while Figure 2 and Figure 4, which was created from Figure 2, are arranged vertically with the time axes aligned and placed on the right. [Figure 6]1 is a schematic diagram of an information processing device for implementing the method of the present invention. [Figure 7] This is an explanatory diagram showing the contribution of each item to the Mahalanobis distance at each date and time, accumulated in chronological order. DETAILED DESCRIPTION OF THE INVENTION

[0031] <Embodiment Example 1> Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings. However, the components described in this embodiment are merely examples and are not intended to limit the scope of the present invention to those components. Note that, although a portion of Figures 1 and 5 described below shows a contribution stack diagram in which the contribution of each item is stacked and displayed, for the sake of convenience, the contribution is shown in increments of 300, and does not necessarily show the maximum contribution of each item.

[0032] <Mahalanobis distance D M Calculation of> The "MT method" is a multivariate analysis method that uses the "Mahalanobis distance," a distance that takes into account the correlation between multiple variables, to distinguish between normal and abnormal states of a target. First, a correlation coefficient matrix relating to the correlation coefficients between each item of the reference space data is created, and then the "inverse matrix of the correlation coefficient matrix" of the reference space is used to calculate the Mahalanobis distance D M This allows us to understand the state change of the target space relative to the reference space.

[0033] In addition, the Mahalanobis distance D M For the calculation method of , see, for example, the technical literature "Try and Learn! Introduction to Quality Engineering MT System Analysis Method" (Nikkan Kogyo Shimbun, first published May 30, 2012), and the Mahalanobis distance D using complex number data. M The calculation method of is known based on, for example, the publication of the patent document "Method for analyzing and diagnosing the state of a system having multiple variables, and information processing device using said method" (Patent Application No. 2020-045873), and therefore will not be described here.

[0034] In order for the MT method to be effective, it is essential to "set the reference space" (i.e., how to select data from normal conditions to detect signs of accidents or abnormalities). Setting the reference space involves selecting the period (time axis) and data items (items of various data being measured and acquired) of the data to be targeted.

[0035] <Calculating the contribution of each item> Mahalanobis distance D M When an abnormality is detected, it is possible to determine which item has the strongest influence and which item is the closest to the Mahalanobis distance D M The "degree of contribution," which indicates how much a product contributes to the increase in the number of products, is calculated using the following procedure.

[0036] The Mahalanobis distance D is the Nth item obtained from the system. M Next, remove one item and calculate the Mahalanobis distance D M Then, the Mahalanobis distance D calculated for N items is calculated. M " and "The Mahalanobis distance D for N-1 items, excluding one item M This difference is the "degree of contribution" for the item in question (the one item removed). By performing these steps for each item, the "degree of contribution" for each item is calculated. Note that "N" does not indicate a specific number, but is calculated by the Mahalanobis distance D M indicates the number of items used to calculate

[0037] <Displaying 100% of the contribution stack> For each date and time, the calculated contribution for each item is divided by the total contribution for all items to calculate the contribution ratio for each item. A color is also assigned to each item. Then, for each date and time, the contribution for each item is stacked in the assigned color based on the calculated contribution ratio and displayed as 100%. In more detail, for example, all items are stacked in the assigned color along the vertical axis to form a single bar graph showing 100%, and this single bar graph is connected and displayed on the horizontal axis in date and time order. By proceeding with these steps (procedures), the contribution stack is displayed as 100%.

[0038] Figure 1 shows a conventional "contribution stack diagram" that stacks the contribution of each item and displays it in chronological order. The vertical axis of this contribution stack diagram is the contribution value, and the horizontal axis is the date and time. On the other hand, Figure 2 shows a 100% display diagram of the contribution stack, which displays the contribution stack as a percentage.

[0039] In other words, the "100% Contribution Stack Display" clearly displays the ratio and trend of contribution by date and time, which in the conventional "Contribution Stack Display" was impossible to see because the values ​​were small and could not be confirmed unless enlarged.

[0040] This allows you to check the change in the ratio of each item's contribution, and as a result, you can grasp fluctuations and signs of system balance. Furthermore, you can also set a rough idea of ​​how to divide the reference space. In other words, the 100% display chart is a basis for determining the creation of an appropriate reference space. Furthermore, you can check the trends of major items in areas with small contributions.

[0041] <Predictive item extraction display> Each item is arranged on the vertical axis, and the time series change in the magnitude of each item's contribution is color-coded according to the magnitude of the contribution and displayed on the horizontal axis. Specifically, for example, the time series change in the magnitude of each item's contribution is colored using a gradient, for example, with small values ​​in blue, large values ​​in red, and values ​​in between in green, and displayed as a horizontally long box. The boxes for each item are arranged vertically with the same area, and the differences between items at the same time are displayed so that they can be visually seen by the color change of the vertical line. By proceeding through these steps (procedures), items that indicate signs of abnormality are extracted and displayed from the contributions. Figure 3 shows a sign item extraction diagram created from the contribution stack diagram.

[0042] For each date and time, the calculated contribution rate for each item is divided by the total contribution rate for all items to calculate the contribution rate for each item. Then, each item is arranged on the vertical axis, and the time series changes in the calculated "contribution rate for each item" are color-coded according to the magnitude of the contribution rate and displayed on the horizontal axis. For example, the time series changes in the "contribution rate for each item" are displayed on the horizontal axis as horizontal boxes, using a color gradient, e.g., blue for small values, red for large values, and green for values ​​in between. The "boxes for each item" are arranged vertically with the same area, and the differences between items at the same time are displayed so that they can be visually seen by the color change of the vertical line. By following these steps, items that indicate signs of an accident or abnormality are extracted and displayed. Figure 4 shows a predictive item extraction diagram created from the contribution rate stack 100% display diagram.

[0043] Figure 3 shows the overall changes over time and the shift in the system balance. Figure 4 provides predictive information for periods with low contribution on the left side. For example, in the first half of the time axis (left side), the area with high contribution (time x item) shifts to the bottom right, and the system appears to have stopped at the thick vertical red line on 6 / 10.

[0044] By reading the items from the right edge, you can see the transitions in items such as "wood chip supply amount," "fluidized bed temperature," "boiler water pH," "exhaust pressure," and "boiler water conductivity." The transitions up to the system shutdown are highlighted, allowing you to see the detailed changes in items with high contributions.

[0045] Moreover, by displaying the strength of the contribution with color, unnecessary items (for example, items with a small contribution, which are displayed in blue in Figures 3 and 4) can be easily identified. Also, it is possible to grasp the transition of items with a large contribution.

[0046] In this way, by extracting predictive items and displaying them as a predictive item extraction diagram, the information related to each item is integrated and displayed, making it possible to visualize and visualize awareness and anomalies (contradictions). Furthermore, by displaying each item grouped by function, the information catalytic effect is promoted (= the integrated display of information promotes awareness and anomalies), inducing visual awareness (contradictions) when assembling abnormality factors and their mechanisms and logic. It also promotes a bird's-eye view of the situation, as if viewing an actual system. Furthermore, this bird's-eye view induces the discovery of not only individual issues, but also issues and inconsistencies that span systems. It also leads to the detection of unexpected abnormalities and the discovery of new phenomena.

[0047] Next, the differences between the "predictive item extraction diagram created from the contribution stack diagram" and the "predictive item extraction diagram created from the contribution stack 100% display diagram" will be specifically explained using Figures 3 to 5. Note that Figure 5 is a comparison diagram in which Figure 1 and Figure 3 created from the same diagram are arranged vertically with the time axes aligned and placed on the left, and Figure 4 created from Figure 2 and the same diagram are arranged vertically with the time axes aligned and placed on the right.

[0048] Looking at the left side of Figure 5, each item is color-coded in ascending order of contribution (blue → green → red), making it possible to grasp the overall balance state of the system. Also, the starting point of change for each item can be seen from the points where the color changes from blue to green. Furthermore, the transitions between items can be seen. It is also easy to grasp changes in items that have something in common, such as boiler water pH and boiler water conductivity, shaft movement and shaft vibration (circles 2 and 4 in the diagram on the left side of Figure 5).

[0049] Looking at the right side of Figure 5, items with large contributions, such as wood chip supply amount, boiler water pH, boiler water conductivity, and shaft movement, are clearly displayed in red. Additionally, the starting point of each item's change can be seen from the points where the color changes from "blue to green." Furthermore, items with commonalities, such as boiler water pH and boiler water conductivity, shaft movement and shaft vibration, change from "blue to red" or "green to red," making it possible to see the starting point of each change. For example, the wood chip supply amount and boiler water pH are clearly shown to have changed from "blue to red" around April, indicating the occurrence of warning signs (circles 1 and 2 in the diagram on the right side of Figure 5).

[0050] Comparing Figures 3 and 4, the occurrence periods of items with high contributions are consistent in both figures.

[0051] Furthermore, "Figure 3," a predictive item extraction diagram created from the contribution stack diagram, clearly shows the balance changes for all items. "Figure 4," a predictive item extraction diagram created from the contribution stack 100% display diagram, clearly shows the signs (period, items) of accidents and abnormalities. In detail, in the predictive item extraction diagram created from the contribution stack diagram, where the values ​​of items that are likely to show signs are low, the changes are not visible because they are crushed. However, in the predictive item extraction diagram created from the contribution stack 100% display, each item is enlarged, so the changes become visible.

[0052] Therefore, the 100% contribution stack display diagram and the predictive item extraction diagram (2 types) are not necessarily used individually, and since each diagram provides unique information, you can display the four diagrams side by side along with the contribution stack diagram and compare them.

[0053] Then, delete the appropriate items, recreate the correlation coefficient matrix with the new item set, and calculate the Mahalanobis distance D M , recalculate the contribution, and compare the four figures. If you feel that the deletion of the item is inappropriate, return the item set to its original state. Alternatively, if you feel that further item deletion is necessary, delete the item, recreate the correlation coefficient matrix with the new item set, and calculate the Mahalanobis distance D from the inverse matrix. M , recalculate the contribution, and compare the four figures. That is, the procedure of "compare four figures" → "delete item" → "calculate contribution" → "compare four figures" can be repeated until an appropriate reference space is obtained.

[0054] <Configuration of information processing device 1> Next, an information processing device 1 for implementing a system balance fluctuation extraction method that can accurately extract fluctuations in system balance and their signs will be described.

[0055] This information processing device 1 calculates the Mahalanobis distance D M The contribution of each item is calculated, and fluctuations in the system balance and their signs are displayed.

[0056] Next, the hardware configuration of the information processing device 1 will be described with reference to Fig. 6. Fig. 6 is a conceptual diagram that schematically illustrates the hardware of the information processing device 1.

[0057] In Figure 6, the control means 11 is realized by, for example, a CPU, and executes application programs, operating systems (OS), control programs, etc. stored on a hard disk (HD) included in the storage means 12 described later, and controls the temporary storage of information, files, etc. necessary for executing the programs in the RAM included in the storage means 12.

[0058] In particular, the control means 11 calculates the Mahalanobis distance D MThat is, the control means 11 creates a correlation coefficient matrix relating to the correlation coefficients between the items of the reference space data, and calculates the Mahalanobis distance D M Calculate.

[0059] In addition, the control means 11 calculates the contribution of each item. That is, the control means 11 calculates the Mahalanobis distance D M Next, remove one item and calculate the Mahalanobis distance D M Then, the Mahalanobis distance D calculated for all N items is calculated. M " and "The Mahalanobis distance D for N-1 items, excluding one item M This difference is the "degree of contribution" for that item. By performing these steps for each item, the "degree of contribution" for each item is calculated.

[0060] Next, the control means 11 calculates the contribution rate of each item by dividing the calculated contribution rate for each item by the total value of the contribution rates for all items for each date and time. It also assigns a color to each item. Then, for each date and time, the contribution rate of each item is stacked in the assigned color based on the calculated contribution rate and displayed as 100%. In more detail, for example, a single bar graph showing 100% of all items stacked in the assigned color along the vertical axis is displayed, and the single bar graph is connected along the horizontal axis in date and time order and displayed on the display means 14.

[0061] Alternatively, the control means 11 arranges each item on the vertical axis, and the time series change in the magnitude of the contribution of each item is color-coded according to the magnitude of the contribution, arranged on the horizontal axis, and displayed on the display means 14. In more detail, for example, the time series change in the magnitude of the contribution of each item is colored using a gradation, for example, with small values ​​in blue, large values ​​in red, and values ​​in between in green, and arranged on the horizontal axis as a horizontally long box, and the boxes for each item are arranged vertically with the same area, and displayed on the display means 14 so that the differences between items at the same time can be visually seen by the color change of the vertical line.

[0062] Alternatively, the control means 11 calculates the contribution rate of each item by dividing the calculated contribution rate for each item by the total value of the contribution rates for all items for each date and time. Then, the control means 11 arranges each item on the vertical axis, and the time-series changes in the calculated "contribution rate of each item" are color-coded according to the magnitude of the contribution rate and arranged on the horizontal axis and displayed on the display means 14. In more detail, for example, the time-series changes in the "contribution rate of each item" are colored using a gradation, for example, with small values ​​in blue, large values ​​in red, and values ​​in between in green, and arranged on the horizontal axis as horizontally long boxes, and the "boxes for each item" are arranged vertically with the same area, and the differences between items at the same time are displayed on the display means 14 so that they can be visually seen by the color change of the vertical line.

[0063] The storage means 12 is for temporarily storing various information, and includes a RAM that functions as the main memory and work area of ​​the control means 11, and a ROM that stores programs such as a basic I / O program and various information used in basic processing. It also includes a HD that functions as a large-capacity memory. The storage means 12 also includes a storage device that stores the Mahalanobis distance D M , contribution rate, 100% stack chart, predictor item extraction chart, etc. may be stored.

[0064] The input means 13 receives input of information and commands from the user to the information processing device 1. For example, it is a keyboard, a touch panel, or a button. Note that the user may operate the control means 11 through the input means 13. In more detail, for example, the control means 11 may be configured to input the Mahalanobis distance D M The degree of contribution is calculated. For each date and time, the calculated degree of contribution for each item is divided by the total value of the degrees of contribution for all items to calculate the percentage of the degree of contribution for each item. A color is also assigned to each item. Then, for each date and time, the degree of contribution for each item is stacked in the assigned color based on the calculated percentage of the degree of contribution and displayed as 100%. Specifically, for example, all items are stacked in the assigned color along the vertical axis to form a single bar graph showing 100%, and the single bar graph is connected along the horizontal axis in date and time order and displayed on the display means 14.

[0065] Furthermore, the control means 11 arranges each item on the vertical axis, and the time series change in the magnitude of the contribution of each item is color-coded according to the magnitude of the contribution, arranged on the horizontal axis, and displayed on the display means 14. Specifically, for example, the time series change in the magnitude of the contribution of each item is colored using a gradation, for example, with small values ​​in blue, large values ​​in red, and values ​​in between in green, and arranged on the horizontal axis as a horizontally long box, and the boxes for each item are arranged vertically with the same area, and displayed on the display means 14 so that the difference between items at the same time can be visually seen by the color change of the vertical line.

[0066] The control means 11 also calculates the contribution rate of each item by dividing the calculated contribution rate for each item by the total value of the contribution rates for all items for each date and time. Then, the control means 11 arranges each item on the vertical axis, and the time-series changes in the calculated "contribution rate of each item" are color-coded according to the magnitude of the contribution rate and arranged on the horizontal axis and displayed on the display means 14. Specifically, for example, the time-series changes in the "contribution rate of each item" are colored using a gradation, for example, with small values ​​in blue, large values ​​in red, and values ​​in between in green, and arranged on the horizontal axis to form horizontally long boxes, and the "boxes for each item" are arranged vertically with the same area, and displayed on the display means 14 so that the differences between items at the same time can be visually seen by the color change of the vertical line.

[0067] It also causes the control means 11 to delete an item and restore an item to its original state.

[0068] The display means 14 is, for example, a liquid crystal display, an organic EL display, or a dot matrix display, and displays commands input from the input means 13 and the corresponding response output from the information processing device 1. In particular, the display means 14 displays a contribution stack diagram, a 100% display diagram of the same, a predictive item extraction diagram, a 100% extraction diagram of the same, etc.

[0069] The bus 16 controls the flow of data within the information processing device 1. The communication means 15 is an interface (I / F), and the information processing device 1 is connected to external devices via this communication means 15.

[0070] It should be noted that software that realizes the same functions as the above devices can be used in place of the hardware devices.

[0071] In addition, in the first embodiment, the program and related data according to this first embodiment can be directly loaded into the storage means 12 such as RAM and executed, but each time the program according to this first embodiment is run, it may be loaded into the storage means 12 such as HD in which the program is already installed. Also, the program according to this first embodiment can be stored in the storage means 12 such as ROM, configured to form part of the memory map, and executed directly by the control means 11 such as CPU. [Explanation of symbols]

[0072] 1: information processing device, 11: control means, 12: storage means, 13: input means, 14: display means, 15: communication means, 16: bus

Claims

1. A method for displaying fluctuations in system balance and their signs, Calculating the contribution using N items of time series data obtained from the system; A step of calculating a ratio of the contribution degree of each item by dividing the calculated contribution degree of each item by the total value of the contribution degrees of all items for each date and time; assigning a color to each of the percentage-calculated items; A method comprising a step of displaying the contribution of each item for each date and time as a percentage by stacking it in the assigned color based on the calculated contribution rate.

2. In a method for displaying fluctuations in system balance and their signs, Calculating the contribution using N items of time series data obtained from the system; A method characterized by comprising a step of arranging each item on a vertical axis, color-coding the time series change in the magnitude of the contribution of each item according to the magnitude of the contribution, and arranging and displaying the change on a horizontal axis.

3. A step of calculating a ratio of the contribution degree of each item for each date and time by dividing the calculated contribution degree of each item by the total value of the contribution degrees of all items; The method according to claim 2, further comprising a step of arranging each item on a vertical axis, color-coding the time series changes in the calculated "percentage of contribution of each item" according to the magnitude of the percentage of contribution, and displaying them on a horizontal axis.

4. An information processing device used in a method for displaying fluctuations in system balance and signs thereof, The information processing device includes: Calculate the contribution using N items of time series data obtained from the system, For each date and time, the calculated contribution rate for each item is divided by the total value of the contribution rates for all items to calculate the contribution rate for each item; Assign a color to each item whose percentage was calculated, An information processing device characterized by having a control means for displaying the contribution of each item for each date and time as 100% by stacking it in the assigned color based on the calculated contribution rate.

5. A program for operating an information processing device having a control means, which is used in a method for extracting fluctuations in system balance and signs thereof, The program causes the control means of the information processing device to: Calculate the contribution using N items of time series data obtained from the system, For each date and time, the calculated contribution degree for each item is divided by the total value of the contribution degrees for all items to calculate the proportion of the contribution degree for each item; Assign a color to each item whose percentage was calculated, A program characterized in that the contribution of each item for each date and time is displayed as 100% by stacking it in the assigned color based on the calculated contribution rate.

6. An information processing device used in a method for extracting fluctuations in system balance and signs thereof, The information processing device includes: Calculate the contribution using N items of time series data obtained from the system, An information processing device characterized by having a control means for arranging each item on a vertical axis, color-coding the time series change in the magnitude of the contribution of each item according to the magnitude of the contribution, and displaying them on a horizontal axis.

7. A program for operating an information processing device having a control means, which is used in a method for extracting fluctuations in system balance and signs thereof, The program causes the control means of the information processing device to: Calculate the contribution using N items of time series data obtained from the system, A program characterized by arranging each item on a vertical axis, color-coding the time series change in the magnitude of the contribution of each item according to the magnitude of the contribution, and displaying the same on a horizontal axis.

8. The control means divides the calculated contribution degree for each item by the total value of the contribution degrees for all items for each date and time to calculate a ratio of the contribution degree for each item for each date and time; 7. The information processing device according to claim 6, characterized in that each item is arranged on a vertical axis, and the time series changes in the calculated "proportion of contribution of each item" are color-coded according to the magnitude of the proportion of contribution and displayed on a horizontal axis.

9. The program causes the control means of the information processing device to For each date and time, the calculated contribution degree for each item is divided by the total value of the contribution degrees for all items to calculate the proportion of the contribution degree for each item for each date and time; The program described in claim 7, characterized in that each item is arranged on the vertical axis, and the time series changes in the calculated ``contribution ratio of each item'' are color-coded according to the magnitude of the contribution ratio and displayed on the horizontal axis.

Images