Flame monitoring device, flame monitoring program, and flame monitoring method

Abstract

To provide a flame strength monitor by which a user can properly grasp a malfunction of a combustion device.SOLUTION: A flame strength monitor comprises an acquisition part 21A for acquiring the newest time series data indicating a continuous secular change of the strength of a flame of a burner; and a comparison part 21B for creating malfunction score time series data indicating a secular change of a difference of the strength of the flame by comparing the newest time series data and reference time series data being references which are prepared as time series data indicating the continuous secular change of the strength of the flame of the burner. The reference time series data include the aggregation of time series data indicating the continuous secular change of the strength of the flame of the burner in each of a plurality of times of combustion sequences which are executed before the occurrence of a malfunction at a combustion device, and before the combustion sequence at which the newest time series data are obtained.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present invention relates to a flame monitoring device, a flame monitoring program, and a flame monitoring method for monitoring the activity of a burner flame. [Background technology]

[0002] Patent Document 1 discloses a technology that monitors the burner flame activity (here, flame strength) for each of a number of subsequences ("pilot trial," "pilot only," "main trial," and "main stable"), and determines that the combustion device is in a malfunction if the flame activity is within a normal range defined for each subsequence but deviates from the normal state. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2019-60573 A Summary of the Invention [Problem to be solved by the invention]

[0004] The technology described in the above Patent Document 1 is also useful, but in this technology, whether or not the combustion device is malfunctioning is determined based on the flame activity at a certain timing. Therefore, with the above technology, it is difficult for the user to grasp the malfunction of the combustion device when the malfunction occurs at a timing other than the certain timing.

[0005] An object of the present invention is to provide a flame monitoring device, a flame monitoring program, and a flame monitoring method that allow a user to appropriately grasp malfunctions in a combustion device. [Means for solving the problem]

[0006] In order to solve the above problem, a flame monitoring device according to a first aspect of the present invention comprises an acquisition unit that acquires first time series data indicating a continuous change over time in the activity of a burner flame, and a comparison unit that compares the flame activity of the first time series data with the flame activity of second time series data that serves as a reference and is prepared as time series data indicating the continuous change over time in the activity of the burner flame, with corresponding data in time series, to generate third time series data indicating the continuous change over time in the difference in the flame activity.

[0007] The burner flame may be ignited and controlled by a combustion sequence executed by a combustion device equipped with the burner, and the second time series data may include a collection of time series data showing continuous changes over time in activity of the burner flame in each of a plurality of combustion sequences executed before a malfunction occurs in the combustion device and before the combustion sequence in which the first time series data was obtained.

[0008] The comparison unit may be configured to output a graph correlating a graph of the flame activity change over time in the first time series data with a graph of the difference change over time in the third time series data.

[0009] The comparison unit may compare the flame activity between the first time-series data and the second time-series data by kernel density estimation, and the difference is a numerical value representing a kernel density estimate.

[0010] The flame of the burner may be ignited and controlled by a combustion sequence executed by a combustion device equipped with the burner, and the comparison unit may be configured to estimate the presence or absence of a malfunction of the combustion device and, if a malfunction is present, the type of the malfunction, based on the third time-series data, and output the estimation result.

[0011] A flame monitoring program according to a second aspect of the present invention causes a computer to execute an acquisition step of acquiring first time series data indicating continuous time changes in the activity of a burner flame, and a comparison step of comparing the flame activity of the first time series data with the flame activity of second time series data that serves as a reference and is prepared as time series data indicating continuous time changes in the activity of the burner flame, with corresponding data in time series to generate third time series data indicating continuous time changes in the difference in the flame activity.

[0012] A flame monitoring method according to a third aspect of the present invention includes a comparison step of comparing the flame activity of first time series data indicating a continuous change over time in the flame activity of a burner with the flame activity of second time series data serving as a reference and prepared as time series data indicating a continuous change over time in the flame activity of the burner, with corresponding data in time series to generate third time series data indicating a continuous change over time in the difference in the flame activity. Effect of the Invention

[0013] According to the present invention, a user can appropriately recognize malfunctions in a combustion device. [Brief description of the drawings]

[0014] [Figure 1] FIG. 1 is a configuration diagram of a combustion system having a flame monitoring device according to a first embodiment of the present invention. [Diagram 2] FIG. 2 is a hardware configuration diagram of the flame monitoring device. [Diagram 3] FIG. 3 is a configuration diagram of a flame monitoring device. [Figure 4] FIG. 4 is a flowchart of the time-series data storage process executed by the acquisition unit. [Diagram 5] FIG. 5 is a flowchart of the reference time series data processing executed by the comparison unit. [Figure 6] FIG. 6 is a flowchart of the comparison process executed by the comparison unit. [Figure 7]FIG. 7 is a diagram showing a graph of the reference time series data. [Figure 8] FIG. 8 is a diagram showing graphs of the most recent time-series data and the malfunction score time-series data. [Figure 9] FIG. 9 is a diagram showing graphs of the most recent time-series data and the malfunction score time-series data. [Figure 10] FIG. 10 is a diagram showing graphs of the most recent time-series data and the ailment score time-series data. [Figure 11] FIG. 11 is a diagram showing graphs of the most recent time-series data and the ailment score time-series data. [Figure 12] FIG. 12 is a flowchart of an estimation result output process executed by a comparison unit according to a modified example. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] Hereinafter, an embodiment of the present invention and its modified examples will be described with reference to the drawings.

[0016] (Embodiment) As shown in Fig. 1, a flame monitoring device 20 according to one embodiment of the present invention is used in a combustion system 10. The flame monitoring device 20 monitors the flame activity of a burner provided in a combustion device 40 (described below) of the combustion system 10, more specifically, a main burner 42 and a pilot burner 43, to detect a malfunction of the combustion device 30. "Malfunction" refers to an abnormality that allows the combustion device 30 to operate to the extent that it is not yet inoperable, that is, a minor abnormality.

[0017] In addition to the flame monitoring device 20, the combustion system 10 includes a combustion device 30 that performs combustion, and a combustion control device 70 that controls the combustion device 30. Below, the combustion device 30 and the combustion control device 70 will be explained first, and then the flame monitoring device 20 will be explained.

[0018] The combustion device 30 includes a combustion appliance 40, a fuel supply system 50, an air supply system 60, a control motor M, a high opening position sensor HS, and a low opening position sensor LS.

[0019] The combustion equipment 40 burns the fuel gas in the combustion chamber R. The combustion equipment 40 includes a combustion furnace 41 that forms the combustion chamber R, a main burner 42 that burns the fuel gas to heat the inside of the combustion chamber R, a pilot burner 43 that burns the fuel gas to ignite the main burner 42, and an ignition device (igniter) 44 that ignites the pilot burner 43. The pilot burner 43 ignited by the ignition device 44 is used to ignite the main burner 42. Therefore, the ignition device 44 can also be said to be a device used when igniting the main burner 42. The ignition device 44 includes a spark generating unit 44A such as a discharge electrode.

[0020] The combustion equipment 40 further includes a flame detector 45 that detects the flame activity of the main burner 42 and the pilot burner 43, and a temperature sensor 46 that detects the temperature inside the combustion chamber R. The flame activity is a degree that indicates how actively the flame is generated, and in this case, it is the flame intensity.

[0021] The flame detector 45 includes a discharge tube 45A including two electrodes and a package that houses the two electrodes together with a predetermined gas. The flame detector 45 further includes a window 45B that transmits electromagnetic waves (here, ultraviolet rays) radiated from the flame of the main burner 42 or the pilot burner 43 and guides them to the discharge tube 45A. In the discharge tube 45A, when a voltage is applied between the two electrodes and the electromagnetic waves radiated from the flame of the main burner 42 or the pilot burner 43 are incident on the package, a discharge occurs between the two electrodes. This discharge causes an electric signal to be output from the flame detector 45. In this way, the flame detector 45 converts the intensity of the electromagnetic waves radiated from the flame of the main burner 42 or the pilot burner 43, that is, the activity of the flame, into an electric signal and outputs it. Here, the output electric signal is a voltage signal whose voltage value changes according to the activity of the flame, but the electric signal may be a current signal whose current value changes according to the activity of the flame. The voltage signal is also called the flame voltage VF, and the current signal is also called the flame current IF. The voltage value of the flame voltage VF and the current value of the flame current IF indicate the activity of the flame. In the following description, the flame voltage VF and its voltage value can be replaced with the flame current IF and its current value.

[0022] Returning to Fig. 1, the fuel supply system 50 supplies fuel gas from the outside to the combustion appliance 40. The fuel supply system 50 includes a fuel flow path 51 through which the fuel gas supplied to the combustion appliance 40 flows. The fuel flow path 51 includes a main flow path 51A through which the fuel gas is supplied from the outside, and a first flow path 51B and a second flow path 51C branched from the main flow path 51A. The first flow path 51B is connected to the main burner 42, and the second flow path 51C is connected to the pilot burner 43.

[0023] The fuel supply system 50 further includes a gas pressure switch 52 provided in the main flow path 51A, main valves 54A and 54B which are safety shutoff valves provided in the first flow path 51B, and pilot valves 54C and 54D which are safety shutoff valves provided in the second flow path 51C. The gas pressure switch 52 is turned ON when the pressure of the fuel gas supplied from the outside is equal to or higher than a predetermined pressure. The main valves 54A and 54B open and close the first flow path 51B. The pilot valves 54C and 54D open and close the second flow path 51C. The fuel supply system 50 also includes a damper 55 for adjusting the fuel flow rate provided in the main flow path 51A.

[0024] The air supply system 60 supplies air to the combustion appliance 40. The air supply system 60 includes an air flow path 61 through which air to be supplied to the combustion appliance 40 flows, and a blower 62 that supplies air to the air flow path 61. The air flow path 61 includes a main flow path 61A connected to the blower 62, and a first flow path 61B and a second flow path 61C branched from the main flow path 61A. The first flow path 61B is connected to the main burner 42. The second flow path 61C is connected to the pilot burner 43.

[0025] The air supply system 60 further includes an air pressure switch 63 that is provided in the main flow path 61A and turns on when the air pressure from the blower 62 is equal to or higher than a predetermined air pressure. The air supply system 60 further includes a damper 65 for adjusting the air flow rate that is provided in the main flow path 61A.

[0026] The dampers 55 and 65 for adjusting the fuel or air flow rate are driven by a control motor M in a linkage to control the opening of the fuel passage 51 and the air passage 61. Driven by the linkage, the opening of each of the dampers 55 and 65 is controlled so that the air-fuel ratio, which is the ratio of fuel to air supplied to the main burner 42, is maintained at a desired ratio (a ratio within an ideal range for combustion). The amount of fuel and air supplied to the main burner 42 is adjusted by the opening of each of the dampers 55 and 65, and thereby the flame activity of each burner is adjusted, which in turn controls the heating temperature for heating the combustion chamber R.

[0027] The high opening position sensor HS detects when the dampers 55 and 65 reach a predetermined high opening position. The low opening position sensor LS detects when the dampers 55 and 65 reach a predetermined low opening position. Each of the sensors HS and LS turns ON when the dampers 55 and 65 reach the high opening position or the low opening position.

[0028] The combustion control device 70 is composed of various computers such as a PLC (Programmable Logic Controller) and a personal computer. The combustion control device 70 is also called a burner controller. The combustion control device 70 monitors the states (ON / OFF) of the gas pressure switch 52, the wind pressure switch 63, the high opening position sensor HS, and the low opening position sensor LS. Various signals are input to the combustion control device 70 from the flame detector 45, the temperature sensor 46, and the like. The combustion control device 70 controls the ignition device 44, the main valves 54A and 54B, the pilot valves 54C and 54D, the blower 62, and the dampers 55 and 65, and the like, based on the monitoring results and various signals.

[0029] The combustion control device 70 controls the operation of the combustion device 30 (ignition and flame control of burners including the main burner 42 and pilot burner 43) according to a predetermined combustion sequence. This combustion sequence includes subsequences such as "start check", "pre-purge", "waiting for ignition", "pilot ignition", "pilot only", "main ignition", "main stability", and "steady combustion". Although details of these subsequences are omitted, in "pilot ignition", the pilot valves 54C and 54D are opened and the ignition device 44 operates to ignite the pilot burner 43. In "pilot only", the operation of the ignition device 44 ends. In "main ignition", the main valves 54A and 54B are opened, and the main burner 42 is ignited using the fire of the pilot burner 43 as a pilot flame. After that, in "main stability", the flame of the main burner 42 is stabilized. In the subsequent "steady combustion", the control motor M is feedback-controlled so that the temperature in the combustion chamber R indicated by the detection signal supplied from the temperature sensor 46 becomes a target value, using the temperature as a feedback value. As the feedback control, for example, proportional control (P control), proportional-integral control (PI control), or proportional-integral-derivative control (PID control) is adopted. Here, the pilot burner 43 is kept ignited in the "main stable" and "steady combustion", but the pilot burner 43 may be extinguished during these periods. When the pilot burner 43 is kept ignited, the flame detector 45 may include a flame detector for the main burner 42 and a flame detector for the pilot burner 43.

[0030] The combustion control device 70 performs analog-to-digital conversion on the flame voltage VF output from the flame detector 45 in the combustion sequence at a predetermined sampling rate, generates time-series data indicating the time change in flame activity indicated by the voltage value of the flame voltage VF, and outputs the data to the flame monitoring device 20. The flame voltage VF here indicates the sum of the flame activity of the main burner 42 and the flame activity of the pilot burner 43.

[0031] Next, the flame monitoring device 20 will be described. The flame monitoring device 20 performs processing to monitor the flame activity of the main burner 42 and the pilot burner 43 detected by the flame detector 45 based on the time series data output from the combustion control device 70. The flame monitoring device 20 is configured to include various computers such as a personal computer. As shown in FIG. 2, the flame monitoring device 20 includes a processor 21 such as a CPU (Central Processing Unit), a RAM (Random Access Memory) 22 that functions as the main memory of the processor 21, and a non-volatile storage device 23 that stores a flame monitoring program executed by the processor. The storage device 23 also stores a time series data group and reference time series data, which will be described later. The flame monitoring device 20 further includes a display 24 that displays various screens, which will be described later, an operation device 25 operated by a user, and a communication module 26 for the processor 21 to communicate with the combustion control device 70.

[0032] In this embodiment, the processor 21 executes a flame monitoring program stored in the storage device 23 to operate as an acquisition unit 21A and a comparison unit 21B shown in FIG.

[0033] The acquisition unit 21A communicates with the combustion control device 70 via the communication module 26, acquires time series data output from the combustion control device 70 every time a combustion sequence is executed, and stores the data in the storage device 23. For example, the acquisition unit 21A executes the time series data storage process shown in FIG.

[0034] In the time series data storage process of FIG. 4, the acquiring unit 21A stores the time series data in the Nth storage area (initial value is 0) in the time series data storage area provided in the storage device 23 (step S11). After that, the acquiring unit 21A judges whether N is 9 (step S12), and if N is not 9 (No), adds 1 to N (step S13). If N is 9 (step S12; Yes), the acquiring unit 21A initializes N to 0 (step S14). When 10 pieces of time series data of N=0 to 9 are stored in the storage device 23 by such a series of processes, the time series data obtained thereafter is overwritten starting from N=0. As a result, the latest 10 pieces of time series data are always stored in the storage device 23. A data group of these 10 pieces of time series data is a time series data group.

[0035] The acquiring unit 21A further acquires the reference time series data and stores it in the storage device 23 by starting the reference time series data process shown in FIG. 5 from the initial operation start after the manufacture of the combustion device 30. In the reference statistical data generation process shown in FIG. 5, the acquiring unit 21A monitors each storage area N=0 to 9 of the storage device 23 and waits until time series data is stored in all of them (step S21). When time series data is stored in each storage area N=0 to 9 (step S21; Yes), the acquiring unit 21A reads out the time series data, that is, 10 pieces of time series data, from each of these storage areas (step S22). Furthermore, the acquiring unit 21A stores a set of the read 10 pieces of time series data in another storage area of ​​the storage device 23 as reference time series data (step S22). Note that the number of time series data constituting the reference time series data is not limited to 10 and is arbitrary.

[0036] The reference time series data processing shown in FIG. 5 may be started before an operation test after the manufacture of the combustion device 30, or may be started after the combustion device 30 enters full operation after the operation test. The above-mentioned manufacture includes the combustion device 30 being renewed by repairing, repairing, modifying, or replacing the flame detector 45. The reference time series data processing may be started at any timing when the user wishes to register the reference time series data. In this case, the user issues an instruction to that effect via the operation device 25. The reference time series data processing may be executed when the combustion device 30 is not malfunctioning. In other words, the reference time series data is generated by a combustion sequence executed after the manufacture of the combustion device 30 and before the combustion device 30 malfunctions.

[0037] Returning to Fig. 3, at any timing after the reference time series data is registered in the storage device 23, the comparison unit 21B acquires the latest time series data from the time series data group stored in the storage device 23 at that time, and compares the acquired latest time series data with the reference time series data. Through this comparison, the below-described malfunction score time series data is generated. The above-mentioned any timing may be, for example, a timing when the user operates the operation device 25 in order to check the malfunction score time series data.

[0038] The comparison unit 21B compares the flame activity of the latest time series data acquired by the acquisition unit 21A with the flame activity of the reference time series data that correspond to each other in the time series, and generates malfunction score time series data, which is time series data that shows the continuous change over time in the difference in flame activity (malfunction score).

[0039] The comparison unit 21B performs the comparison etc. by executing, for example, a comparison process shown in Fig. 6. In the comparison process shown in Fig. 6, the comparison unit 21B first compares the latest time series data with the reference time series data using kernel density estimation (step S31).

[0040] The time series data X for n=10 times constituting the reference time series data to be compared in step S31 i(i = 1, 2, 3... n) is the formula for "X1 = X 1,1 ,X 1,2 ,X 1,3 ,···X 1,M ", "X2=X 2,1 ,X 2,2 ,X 2,3 ,···X 2,M ", "X n =X n,1 ,X n,2 ,X n,3 ,···X n,M " The reference time series data is also called reference time series data X. Here, X on the right side i,1 ~X i,M is the flame activity at each time point constituting the time series data, and is arranged in chronological order. M is the number of flame activity data points, that is, the number of time points. If there is a discrepancy in the sampling timing for obtaining the time series data, or if the number of flame activity data points differs, the comparison unit 21B may perform time expansion / contraction using a known technique such as DTW (Dynamic Time Warping) to match the sampling timing and the number of data points.

[0041] The latest time series data is "Y=Y1,Y2,Y3,...Y M ", the comparison unit 21B calculates the flame activity Y at a certain time point m as m The kernel density estimator fm for is calculated using the following formula (1).

number

[0042] In the formula (1), K is a kernel function, and the comparison unit 21B performs the calculation using, for example, the Gaussian formula of the following formula (2).

number

[0043] In the above formula (1), h is an adjustment parameter called a bandwidth. The comparison unit 21B calculates h by multiplying the flame activity X i,mIt may be calculated by a method such as cross-validation, or it may be calculated heuristically using the Scott method or Silverman method. Here, the Scott method using the following formula (3) is used. The following s is X i,m is the sample standard deviation of , q(0.75) and q(0.25) are the 0.75th and 0.25th quantiles of the sample, respectively, and min(a,b) is the function that takes the smallest value of a and b.

number

[0044] Then, for each calculated fm, the malfunction score Em is defined by the following formula (4). The malfunction score Em is a value that becomes larger the more the time series data Y to be processed at a certain point in time m does not follow the distribution of the reference time series data X (the range of possible flame activity of each flame of the multiple time series data constituting the reference time series data). In other words, if the malfunction score Em is large, there is a high possibility that a malfunction has occurred at that point in time. The kernel density estimator fm and the malfunction score Em can also be said to be the difference in flame activity at each point in time of the time series data.

number

[0045] In this manner, in step S31, the comparison unit 21B compares the time series data Y with the reference time series data X1 to X n (n=10) and the flame activity (Y m ,X 1,m ,X 2,m ,X 3,m ,···X n,m ) and obtains the difference in flame activity at each time point (malfunction score Em). The set of these differences at each time point, that is, the time series data showing the change in these differences over time, is hereinafter also referred to as malfunction score time series data E. The comparison unit 21B generates the malfunction score time series data E through the above comparison and acquisition.

[0046] The comparison unit 21B then correlates the malfunction score time series data E and the latest time series data Y, graphs them, outputs them to the display 24, and displays the graphs of each data on the display 24 (step S32). The user can check the graphs displayed on the display 24 to understand whether or not the combustion device 30 is malfunctioning.

[0047] Here, a graph of the reference time series data X is shown in FIG. 7, and examples of graphs of the time series data Y and the illness score time series data E displayed on the display 24 are shown in FIGS.

[0048] 7, three time series data are depicted as representatives of the ten time series data of the reference time series data X. The graph in Fig. 7 is a graph of time series data when no malfunction occurs in the combustion device 30, that is, when the combustion device 30 is in a normal state (non-abnormal state) before the malfunction occurs, and the actual graphs of the ten time series data have similar shapes.

[0049] Fig. 8 shows a graph of the time series data Y and the malfunction score time series data E obtained from the combustion sequence of the combustion device 30 before a malfunction occurs. As shown in Fig. 8, when no malfunction occurs, the malfunction score Em fluctuates close to 0. In particular, even if the flame activity changes significantly from around 0 (before ignition) to around 3.4 (steady combustion), there is no significant change in the malfunction score Em.

[0050] 9 shows a graph when a malfunction occurs in the combustion device 30, causing a high frequency of false discharges (discharges other than those caused by electromagnetic waves) in the flame detector 45. If a threshold for determining the presence or absence of a malfunction is set near the value of the malfunction score Em=10, it can be seen that the malfunction occurs periodically. A user who sees such a graph can understand that there is a malfunction, and that the type of malfunction may be a false discharge (discharges other than those caused by electromagnetic waves) in the flame detector 45.

[0051] FIG. 10 shows a graph when a malfunction occurs in the combustion device 30, in which the window 45B of the flame detector 45 is dirty with soot or the like. For example, as in FIG. 9, if a threshold for determining whether or not there is a malfunction is set near the value of the malfunction score Em=10, it can be seen that the malfunction has continued for a long time. Note that, when graphing, a limiter is provided for the malfunction score Em, and when the upper limit value is exceeded, the value is converted to the upper limit value. A user who sees the graph in FIG. 10 can understand that there is a malfunction, and that the type of malfunction may be dirt on the window 45B.

[0052] 11 shows a graph when the air-fuel ratio, which is the ratio of air to fuel gas supplied to the main burner 42 and the pilot burner 43, is disturbed and falls outside the range expected in advance. As above, if a threshold value for determining the presence or absence of a malfunction is set near the value of the malfunction score Em = 10, a user looking at such a graph can understand that a malfunction is occurring in each period near the start and end of combustion, and can understand that the type of malfunction may be a disturbance in the air-fuel ratio.

[0053] As described above, by checking the malfunction score time series data E, if there is a period in which the malfunction score Em is high, the user can know that the combustion device 30 is malfunctioning. In particular, since the malfunction score time series data E indicates a continuous time change in the malfunction score Em (the difference between the flame activity in the reference time series data X and the flame activity in the time series data Y), the occurrence of a malfunction is confirmed over a certain period of time rather than at a specific timing. In this way, according to this embodiment, the user can properly know whether or not the combustion device 30 is malfunctioning. If the user can properly know the malfunction of the combustion device 30, the user can sense the signs of a breakdown.

[0054] In this embodiment, the malfunction score Em is obtained by comparing the flame activity of each flame in the reference time series data X and the time series data Y that correspond in time series. Therefore, even if a certain threshold is set for the malfunction score Em, the threshold will change over time. Therefore, an optimal threshold for malfunction judgment is set.

[0055] Furthermore, in the above embodiment, a graph associating the graph of the time change of the flame activity of the time series data Y with the graph of the time change of the malfunction score Em of the malfunction score time series data E is output and displayed on the display 24, so that the user can compare and confirm the time changes of the former and the latter, and can properly understand in which subsequence the malfunction occurs (the period of each subsequence may also be displayed on the graph). In particular, in this embodiment, as shown in FIG. 8 etc., the graph of the time series data Y and the graph of the malfunction score time series data E are displayed superimposed on the same plane as the association. This makes it easier to compare the two graphs. The two graphs may be associated with each other as two graphs with corresponding horizontal axes. The association may be such that the two graphs can be compared. The comparison may be performed for each subsequence, or the graphs may be displayed separately for each subsequence.

[0056] The reference time series data X may be, for example, time series data prepared as a reference showing a continuous time change in the activity of the burner flame. For this reason, the reference time series data X may be data obtained from the results of an experiment using a combustion device of the same type as the combustion device 30 and prepared in advance, but in this embodiment, the reference time series data X is prepared based on the flame voltage VF measured in a combustion sequence actually performed in the combustion device 30. This allows the user to obtain the reference time series data X reflecting the individual characteristics of the combustion devices 30 produced in multiple numbers, so that the user can more appropriately grasp the malfunction of the combustion device 30. Note that the burners mentioned above are both the main burner 42 and the pilot burner 43 here, but either one of them may be the target.

[0057] As described above, the comparison between the time series data Y and the reference time series data X is performed by kernel density estimation, so that a highly accurate illness score Em can be obtained. Furthermore, by graphing the illness score Em, which is a numerical value representing the kernel estimator fm, the user can intuitively understand that if the illness score Em is large, the possibility of illness is high. Note that the numerical value to be converted into time series data and graphed is not limited to the illness score Em as long as it is a numerical value representing the kernel estimator fm. The numerical value representing the kernel estimator fm may be the kernel estimator fm itself. In other words, the kernel estimator fm may be converted into time series data and graphed.

[0058] (Modification) The configuration of the above embodiment can be modified as desired. Modifications are shown below as examples. Each modification can be combined with at least a portion of the others.

[0059] (Variation 1) The configuration of the combustion device 30 is arbitrary. For example, the combustion device 30 may be a type having only a main burner 42 without a pilot burner 43.

[0060] (Variation 2) The comparison unit 21B may perform a process of estimating the presence or absence of a malfunction in the combustion device 30 and, if a malfunction occurs, the type of the malfunction, based on the malfunction score time series data E. The comparison unit 21B may perform an estimation result output process shown in FIG. 13, for example, every time new time series data is stored in the storage device 23.

[0061] 12, the comparison unit 21B estimates the presence or absence of a malfunction in the combustion device 30 and, if a malfunction is present, the type of the malfunction based on the malfunction score time-series data E (step S41). If the comparison unit 21B determines that a malfunction is present as a result of the estimation (step S42; Yes), it may output the result of the estimation (step S43).

[0062] As shown in Fig. 9, if a period in which the malfunction score Em exceeds a predetermined threshold occurs periodically and is shorter than a predetermined period, the comparison unit 21B estimates that a malfunction, that is, a false discharge, has occurred in the flame detector 45 of the combustion device 30. As shown in Fig. 10, if a period in which the malfunction score Em exceeds the predetermined threshold occurs periodically and is longer than the predetermined period, the comparison unit 21B estimates that a malfunction, that is, contamination of the window 45B of the flame detector 45 of the combustion device 30, has occurred. As shown in Fig. 11, if the malfunction score Em exceeds the predetermined threshold in each of the periods near the start and end of combustion, the comparison unit 21B estimates that a malfunction, that is, disturbance of the air-fuel ratio, has occurred.

[0063] The comparison unit 21B displays the result of the above estimation on the display 24. Examples of the displayed contents include messages such as "There may be a high frequency of false discharges in the flame detector," "The flame detector window may be dirty," and "The air-fuel ratio is disturbed."

[0064] According to this modification, the user is notified of the presence or absence of a malfunction in the combustion device 30 and, if a malfunction occurs, the type of the malfunction, so that the user can more appropriately grasp the malfunction of the combustion device 30. Note that the malfunction score time-series data E may be appropriately changed to time-series data indicating a continuous time change of the kernel estimator fm.

[0065] (Variation 3) The flame monitoring device 20 may include an acquisition unit 21A that acquires first time series data (time series data Y in the above) showing a continuous time change of the flame activity of a burner (both or at least one of the main burner and the pilot burner, etc.), and a comparison unit 21B that compares the flame activity of the first time series data acquired by the acquisition unit 21A with the flame activity of the second time series data (reference time series data in the above) that is a reference prepared as time series data showing a continuous time change of the flame activity of the burner, with corresponding ones in time series, to generate third time series data (time series data showing a continuous time change of the malfunction score time series data E or the kernel estimator fm in the above) showing a time change of the difference in the flame activity. The specific processing performed by the acquisition unit 21A and the comparison unit 21B is arbitrary. For example, the first time series data and the second time series data may be compared by other methods. According to the above configuration, the user can properly grasp the malfunction of the combustion device 30 by checking the third time-series data.

[0066] (Variation 4) The acquisition unit 21A may acquire time-series data indicating time changes in the air flow rate, fuel gas flow rate, furnace body temperature, pressure, etc., which are appropriately measured by the combustion control device 70 using various sensors provided in the combustion device 30. In this case, for example, when the user specifies a part of concern in the graph of the malfunction score time-series data using the operation device 25, the comparison unit 21B may display time changes in the air flow rate, fuel gas flow rate, furnace body temperature, pressure, etc., for a certain period before and after that part. This allows the user to accurately grasp signs of abnormality in the combustion device 30 and perform appropriate maintenance of the combustion device 30, thereby enabling efficient operation of the combustion device 30.

[0067] (Variation 5) The comparison unit 21B may store a plurality of malfunction score time series data in the storage device 23 and output a graph in which the malfunction score time series data are associated with each other (for example, a graph in which the graphs of the malfunction score time series data are superimposed on the same plane). This allows the user to know whether the malfunction score is temporary or is gradually increasing due to deterioration over time, etc. This allows the user to know the locations where the malfunction score is likely to increase, etc.

[0068] (Variation 6) The flame activity may be indicated by the number of discharges per unit time that occur between the electrodes of the discharge tube 45A of the flame detector 45. When the flame activity is high, for example when the ultraviolet light intensity is high, the number of discharges per unit time increases. In other words, an increase in the number of discharges corresponds to an increase in the voltage value of the flame voltage described above. Note that the number of discharges also increases in the case of false discharges. In such a case, it is advisable to provide the combustion device 70 with a counter that counts the number of discharges and a measuring unit that measures the counting period.

[0069] (Variation 7) The hardware configuration of the flame monitoring device 20 is arbitrary. The acquisition unit 21A and the comparison unit 21C may each be composed of one or more computers or control circuits (such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field-Programmable Gate Array)). At least a part of the units 21A and 21B may be provided in the combustion control device 70. The flame monitoring device 20 may be a server computer, a cloud computer, or the like. The output destination of the malfunction score time series data, the above-mentioned estimation result, and the like may be a display of a user terminal, etc. The output destination of the comparison result may be a printer, a storage medium, a network, another computer, etc. Each device such as the flame monitoring device 20 includes a device in which the components of the device are integrated into one housing, as well as a system in which the components of the device are distributed and accommodated in multiple housings.

[0070] (Variation 8) The flame detector 45 may take in electromagnetic waves from the flame through a window provided in the combustion furnace 41. In this case, if this window provided in the combustion furnace 41 becomes dirty with soot or the like, it will cause the combustion device 30 to malfunction. In other words, the dirt on the window 45B of the flame detector 45 will replace the dirt on the window of the combustion furnace 41. The flame detector 45 may be fitted into the through hole of the combustion furnace 41 from the outside, and in this case, the window of the combustion furnace 41 may consist of the window 45B of the flame detector 45.

[0071] (Flame monitoring method) The processing executed by the flame monitoring device 20 performs a flame monitoring method that compares reference time series data X with most recent time series data Y and generates malfunction score time series data E, but at least a part of the method may be performed by an object or person other than the flame monitoring device 20.

[0072] (Scope of the present invention) Although the present invention has been described above with reference to the embodiments and modifications, the present invention is not limited to the above embodiments and modifications. For example, the present invention includes various modifications to the above embodiments and modifications that can be understood by a person skilled in the art within the scope of the technical concept of the present invention. The configurations listed in the above embodiments and modifications can be appropriately combined within a range without contradiction. [Explanation of symbols]

[0073] 10...combustion system, 20...flame monitoring device, 21...processor, 21A...acquisition unit, 21B...comparison unit, 23...memory device, 25...operation device, 40...combustion equipment, 42...main burner, 43...pilot burner, 44...ignition device, 44A...spark generating unit, 45...flame detector, window...45B, 50...fuel supply system, 60...air supply system, 70...combustion control device.

Claims

1. an acquisition unit that acquires first time series data indicating a continuous change in flame activity of the burner over time; a comparison unit that compares the flame activity of the first time series data with the flame activity of second time series data that serves as a reference and is prepared as time series data that indicates a continuous change over time in the flame activity of the burner, and generates third time series data that indicates a continuous change over time in the difference in the flame activity, the flame of the burner is ignited and controlled by a combustion sequence executed by a combustion device comprising the burner; The comparison unit is configured to estimate the presence or absence of a malfunction of the combustion device and, if a malfunction occurs, the type of the malfunction based on the third time-series data, and output an estimation result. Flame monitoring device.

2. The second time series data includes a set of time series data showing continuous time changes in flame activity of the burner in each of a plurality of combustion sequences executed before a malfunction occurs in the combustion device and before the combustion sequence from which the first time series data was obtained.

2. The flame monitor of claim 1.

3. The comparison unit is configured to output a graph correlating a graph of a time change in the flame activity of the first time series data with a graph of a time change in the difference of the third time series data.

3. A flame monitor according to claim 1 or 2.

4. an acquisition unit that acquires first time series data indicating a continuous change in flame activity of the burner over time; a comparison unit that compares the flame activity of the first time series data with the flame activity of second time series data that serves as a reference and is prepared as time series data that indicates a continuous time change in the flame activity of the burner, by using kernel density estimation to compare corresponding values ​​in time series to generate third time series data that indicates a continuous time change in a kernel density estimate; A flame monitoring device comprising:

5. A flame monitoring program that causes a computer to function as the flame monitoring device according to claim 1 or 4.

6. a comparison step of comparing a flame activity of a first time series data showing a continuous change over time of the flame activity of a burner with a flame activity of a second time series data serving as a reference and prepared as time series data showing a continuous change over time of the flame activity of the burner, with corresponding flames in time series, to generate third time series data showing a continuous change over time of the difference in the flame activity, the flame of the burner is ignited and controlled by a combustion sequence executed by a combustion device comprising the burner; In the comparison step, the presence or absence of a malfunction of the combustion device and, if a malfunction is present, the type of the malfunction are estimated based on the third time-series data. Flame monitoring method.

7. a comparison step of comparing a flame activity of a first time series data showing a continuous time change of the flame activity of a burner with a flame activity of a second time series data serving as a reference and prepared as time series data showing a continuous time change of the flame activity of the burner by using kernel density estimation between corresponding time series to generate third time series data showing a continuous time change of the kernel density estimate; A flame monitoring method comprising:

Images