JPH05172303A - Optimum control device of recovery boiler - Google Patents

Abstract

PURPOSE:To easily select the fuzzy parameter of a control unit controlling combustion on the basis of fuzzy inference by providing a parameter display device displaying the change of the operation parameter of a recovery boiler with the elapse of time and the fuzzy parameter used in fuzzy inference to the control unit. CONSTITUTION:Fuzzy parameters such as the kind of the membership or input/ output gain set to the fuzzy controller of a fuzzy inference part 27 are displayed on the screen of the image display part 54 of a fuzzy parameter adjusting support part 50 and, at the same time, the output of the fuzzy controller and the quantity to be controlled of a recovery boiler are displayed on the same screen with the elapse of time. The set states of the fuzzy parameters at a predetermined time of various operation parameter curves displayed with the elapse of time are read from a memory device 53 to be confirmed. By this constitution, the quantitative analysis or confirmation of the effect of the setting of the parameters on operation parameters becomes easy.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a recovery boiler optimum control device using a fuzzy theory, and more particularly to a recovery boiler equipped with fuzzy parameter adjustment support means for facilitating selection of fuzzy parameters in a fuzzy inference section. It relates to an optimum control device.

[0002]

2. Description of the Related Art A recovery boiler optimum control device will be described with reference to a recovery boiler of a pulp manufacturing plant proposed by the applicant of the present application in Japanese Patent Application No. 2-184608.

Generally, a pulp manufacturing plant has a chip cooking process for cooking chips, which are pieces of wood. In this chip cooking process, cellulose (fiber content) is separated from lignin (fat content) in chips. This is a process of taking out, and for that purpose, it is necessary to use a chemical agent whose main component is NaOH. From the chip cooking process, the dissolved organic components such as lignin and Na ₂ SO ₄ and Na ₂ CO ₃ are dissolved.
A mixed liquid (black liquor) with an inorganic component containing Na as a main component is discharged.

Therefore, in the recovery boiler of this type of pulp manufacturing plant, attention is paid to the above black liquor component, and by using this black liquor as a fuel raw material for the recovery boiler, organic components such as lignin contained in the black liquor are collected. the resulting steam to be supplied to various equipment of the power generation or the plant by heat held, reducing over Na ₂ SO ₄ contained in the black liquor further by utilizing the combustion heat of the black liquor, which mainly Na ₂ S The ingredients are melted in the furnace bottom to obtain the chemical raw material necessary for the chip cooking step.

That is, the recovery boiler has a dual role of burning the black liquor discharged in the pulp manufacturing plant to recover the thermal energy and the chemical raw material, and has a very important meaning in the operation of the pulp manufacturing plant. is doing.

FIG. 6 is a diagram showing a schematic configuration of an operation control system of a recovery boiler realized under such an intention. First,
The black liquor discharged as waste liquid from the chip cooking process is sent to the black liquor heater 1. Here, the black liquor is heated to a predetermined temperature by receiving a heating medium (for example, recovered steam) whose flow rate is adjusted by the black liquor temperature control valve 2. After being controlled, it is injected into the furnace 4 by the black liquor injection gun 3. The black liquor sprayed by the black liquor spray gun 3 is dried and floated by radiant heat in the furnace and drops on the furnace bottom to form a char bed 5. At this time, combustion air is introduced into the bottom of the char bed, the top of the char bed, the top of the char bed, and the like from the blower ports 2L, 2H, 1L, 1H in the furnace 4, and thereby the char bed accumulated in the furnace. 5 and part of the black liquor in the dry floating state burns. As a result, in the char bed 5 and its vicinity, a high-temperature reducing atmosphere is formed by receiving the heat of combustion, Na ₂ SO ₄ mixed in the black liquor is reduced to Na ₂ S, and the raw material for chip digestion is treated with the raw material of chips for cooking. A certain smelt 8 melts and accumulates and flows out from the smelt spout 9 to the outside. The smelt 8 flowing out from the smelt spout 9 is collected in the collection tank 10.

On the other hand, the exhaust gas 11 after combustion in the recovery boiler is discharged to the outside through the heat exchangers 12, 13, ..., At this time, the water supplied to the recovery boiler is each heat exchanger 13. , 12 are heated to a required temperature, finally converted into steam and recovered, and supplied to a power generation turbine (not shown), various facilities in a plant or factory.

In such a recovery boiler, the following points pose a problem.

One of them is that the shape of the mountain of the char bed 5 is a very important requirement for increasing the recovery rate of the smelt 8, but no measures have been taken against it. That is, the recovery rate of the smelt 8 is greatly influenced by the shape of the peaks of the char bed 5 as described later, and the higher the shape of the peaks, the higher the recovery rate.

In the other one, the black liquor floating in the vicinity of just above the char bed 5 rises above the furnace together with the rising gas 11 and adheres to the heat exchangers 12, 13 ,. Due to this, the heat conversion efficiency of the heat exchangers 12, 13, ... Remarkably decreases, dust trouble occurs in the worst case, and a situation in which the boiler must be temporarily stopped occurs. Therefore, it is necessary to reduce the amount of black liquor suspended as much as possible.

Further, another one is that it is necessary to keep SO ₂ , NOx, etc. contained in the gas discharged from the discharge port below a certain standard value from an environmental standpoint.

[0012] However, in the conventional control of the recovery boiler, although it is necessary to control from the above comprehensive point of view, in practice, no concrete measure is taken for it. It is often operated exclusively by the experience of the operator in the field.

Therefore, in order to carry out an optimal operation from a comprehensive viewpoint while considering the original intention of the recovery boiler and environmental problems, to improve recovery efficiency and to ensure stable operation of the recovery boiler, The applicant has proposed an optimum control device for the recovery boiler.

FIG. 7 is a functional block diagram showing an embodiment of the control device for the recovery boiler. In describing the embodiment of the present invention, the same parts as those in FIG. 6 are designated by the same reference numerals and the detailed description thereof will be omitted. In this device, the char bed shape detecting means 21 is provided at an appropriate position on the furnace wall 4 of the recovery boiler.
And a charbed shape recognition means 22 is provided on the output side of the charbed shape detection means 21.

As the char bed shape detecting means 21, for example, a radiation thermometer is provided on the furnace wall portion at a position corresponding to the top of the char bed 5 so that the radiation thermometer can be swung, and the radiation thermometer can be operated according to the temperature obtained by the swing. The shape signal of the char bed 5 is obtained from the level signal, or an image signal representing the position and shape of the char bed 5 is obtained using an infrared camera or the like.

The charbed shape recognizing means 22 receives the output of the charbed shape detecting means 21 and recognizes the degree of conformity to the respective charbed shapes classified as shown in FIG. 9, for example.

The char bed 5 has various shapes, but they are classified into seven in FIG. 9, and FIG. 9A shows an ideal shape (TYPE 0). In the same figure (b), the width of the char bed is wide, the height is high, and the skirt is not wide (TYPE 1). In the figure (c), the width of the char bed is wide, the height is high, and the skirt is wide (TY
PE2). In the same figure (d), the height of the char bed is low and the skirt is not wide (TYPE 3). In the figure (e), the height of the char bed is low and the skirt is wide (TYPE
4). In Fig. 5 (f), the width of the char bed shape is not wide,
The height is high and the base is not wide (TYPE 5). The figure (g) shows the state where the width of the char bed shape is not wide, the height is high, and the skirt is wide (TYPE 6).

The charbed shape recognizing means 22 is specifically composed of a neural network 221 as shown in FIG. That is, the neural network 221 has a multi-layer structure including an input layer, at least one or more intermediate layers, and an output layer, and each layer includes a plurality of neuron elements. Among them, in the input layer, the detection signals corresponding to the number of neuron elements are subdivided from the charbed detecting means 21 and input to each neuron element. For example, if the charbed detection / detection means 21 is an infrared camera, the binarized signal for each pixel is input to the neuron element corresponding to each pixel.
On the other hand, the middle layer is composed of an arbitrary number of neuron elements,
All or a required number of outputs of the input layer are input with desired synapse coupling coefficients, and each neuron element performs a predetermined calculation and outputs.

Next, the output layer is composed of the number of neuron elements corresponding to the type of recognition shape, and the output of all or a required number of neuron elements of the intermediate layer is input with a desired synapse coupling coefficient, Here, a predetermined calculation is performed and a value of 0 to 1 is output as the conformity of the shape classification.
These suitability values are supplied to the fuzzy inference unit 27.

The neural network 221
When offline, the neural network 221 is used to display the recognition result of the charbed on the display unit in advance based on the shape of the mountain of the charbed, and the judgment result of the operator is given to the display content as a teacher signal. While changing the synaptic coupling coefficient, work is performed so that the actual mountain shape and the shape recognition result match. And
The teacher signal is generated so that a human-like recognition result can be obtained online, and the synaptic coupling coefficient can be changed based on the teacher signal.

Further, a suspended matter amount detecting means 24 for detecting the dried suspended matter amount of the black liquor is provided at an appropriate position on the furnace wall 3 of the recovery boiler, and further, a suspended matter amount recognizing means is provided on the output side of the suspended matter amount detecting means 24. 25 are provided. The suspended matter amount detecting means 24 is for detecting the amount of black liquor dried suspended matter in the upper part of the furnace. Specifically, an ITV camera is attached so that the bottom of the heat exchanger 12 can be seen from the furnace wall. The amount of the black liquor dried suspended matter adhered to 12 is detected, or the amount of suspended matter suspended on the top of the char bed is detected by an ITV camera. The floating substance amount recognizing means 25 is the floating substance amount detecting means 2
When the floating amount detection signal for each pixel from 4 is received, it is constituted by a neural network as shown in FIG. 10 like the charbed shape recognizing means 22 and recognizes the amount of floating amount as shown in FIG. The suspended matter amount recognition means 25
The fuzzy inference unit 27 is supplied with three outputs indicating the degree of the dry suspended matter corresponding to each state of “zero”, “small” and “large”. The neural network of the floating material amount recognition means 25 is also set in advance by using a teacher signal based on the judgment result by the operator.

FIG. 11 (a) shows that the amount of black liquor dry suspended matter is not (appropriate), FIG. 11 (b) shows the amount of black liquor dried suspended matter is small, and FIG. 11 (c) shows the amount of black liquor dried suspended matter. It shows a lot of states.

The exhaust gas concentration detecting means 26 detects the exhaust gas concentration of the gas exhaust system. Specifically, S in the gas exhaust system
Gas concentration detector 2 for detecting O ₂ concentration, NOx concentration, etc.
6a is installed and has a function of performing signal processing on the outputs of these detectors and outputting an exhaust gas concentration detection signal representing the SO ₂ concentration and NOx concentration in the exhaust gas to the fuzzy inference unit 27.

The fuzzy reasoning section 27 receives the outputs of the charbed shape recognizing means 22, the suspended matter amount recognizing means 25, and the exhaust gas concentration detecting means 26, obtains an optimum control signal using fuzzy logic, and controls the boiler. Control the system. Therefore, as shown in FIG. 12, the fuzzy inference unit 27 adds the outputs of the four fuzzy controllers and these controllers and adds the control signals Δ1L, Δ1H, Δ2L, Δ2.
And an adder circuit for obtaining H and ΔTB. Each fuzzy controller is roughly divided into an input adjusting unit, a fuzzy inference unit, and an output adjusting unit. In response to a fuzzy parameter setting signal G supplied from the outside, gains, scaling factors, inference rules, and membership of the input and output adjusting units. Select fuzzy parameters such as functions.

Control signals Δ1L, Δ1H, Δ2L and Δ2
H represents the amount of change in the amount of air at each blow port of the recovery boiler 4, and is supplied to the air flow rate setting unit 28. Control signal △
TB is supplied to the black liquor temperature setting unit 29. The black liquor temperature setting unit 29 calculates the current set temperature by adding the temperature change amount ΔTB to the previously set temperature, and controls the opening of the black liquor temperature operation valve 2. By adjusting the black liquor temperature control valve 2, the heating of the black liquor by the black liquor heater 1 is adjusted. Air flow rate setting unit 2
A flow rate signal representing the flow rate of black liquor is supplied to 8 from a black liquor flow rate detection unit 14 that detects the flow rate of black liquor discharged from the nozzle 3.

Based on the flow rate of black liquor, the air flow rate setting unit 28 calculates the total amount of air required to burn the black liquor by means of the air-space ratio. Then, the primary low air flow rate set value 1L is
It is calculated by the sum of the previous value 1L (k-1) and the change Δ1L. Similarly, the primary high air flow rate set value 1H, the secondary low air flow rate set value 2L, and the secondary high air flow rate set value 2H are calculated. Further, allocation is performed so that the sum of the air flow rate set values becomes the total air amount. In this way, the obtained air flow rate set values 1L, 1H, 2L, 2H are supplied to the air supply unit of the recovery boiler 4 to supply the corresponding amount of air to the recovery boiler 4. The fuzzy inference unit 27, the air flow rate setting unit 28, and the black liquor temperature setting unit 29 form a set value changing unit 30.

Next, the operation of the apparatus configured as above will be described. Now, the neural network 221 that constitutes the charbed shape recognition means 22 by detecting the signal for grasping the shape of the charbed 5 by the charbed shape detection means 21 installed on the furnace wall of the recovery boiler 4
Supply to. In each of the neuron elements forming the intermediate layer and the output layer of the neural network 221, a plurality of signals obtained by multiplying a plurality of outputs of the preceding layer by a synapse coupling coefficient are added, and the added value is calculated. By subtracting with a threshold value and converting to an output function (for example, a sigmoid function) and outputting the result, for example, the char bed 5 is offset to the left or right from the neuron element in the output layer, or whether the width of the char bed 5 is wide or narrow. , The shape recognition signals TYPE0 to TYPE6 indicating whether the char bed 5 is high or low, or whether or not the overall shape is good, and the shape of the char bed 5 is classified into the classified shapes shown in FIG. 9 from these signals. Recognize to what extent it fits. At this time, the neural network 221 makes an appropriate judgment while varying the synapse coupling coefficient based on the teacher signal re-learned by the operator when offline as described above, and recognizes the shape based on this judgment result. To send to.

On the other hand, the floating amount recognizing means 25 takes in the floating amount detecting means 24, for example, the image signal of the camera (the image shown in FIG. 11) and makes the following judgment by the neural network 251 to obtain the recognition result. To the fuzzy inference unit 27.

It is appropriate because there is no dry floating amount (FIG. 11).
(A)). The amount of dry suspension is small (Fig. 11 (b)). There is a large amount of dry suspension (Fig. 11 (c)). Further, the fuzzy reasoning section 27 receives the SO ₂ concentration signal and the NOx concentration signal from the exhaust gas concentration detecting means 26.

The fuzzy inference unit 27 is composed of, for example, four fuzzy controllers and an adder group for adding their corresponding outputs, as shown in FIG. As a fuzzy inference method in each fuzzy controller, an appropriate one is selected from several proposed methods and used. Inference Rule Using Charbed Shape Recognition Result The fuzzy controller 1 uses the output result of the neural network for recognizing the shape as the antecedent part of the logical condition, and performs fuzzy inference using the following inference rules to obtain Δ1L, Δ1H. , Δ2L, ΔTB are calculated. R ₁₁ : IF shape is TYPE 0 THEN Δ1L is Z Δ1H is Z Δ2L is Z ΔTB is Z R ₁₂ : IF shape is TYPE 1 TYPE 1 THEN Δ1L is Z Δ1H is P Δ2L is P ΔTB is PS R ₁₃ : IF shape is TYPE2 THEN Δ1L is PS Δ1H is P Δ2L is P ΔTB is PS R ₁₄ : IF shape is TYPE3 THEN Δ1L is Z Δ1H is N Δ2L is N ΔTB is NS R ₁₅ : IF Shape is TYPE4 THEN Δ1L is PS Δ1H is N Δ2L is N ΔTB is NR ₁₆ : IF shape is TYPE5 THEN Δ1L is Z Δ1H is P Δ2L is P ΔTB is PR ₁₇ : IF shape is TYPE6 THEN △ 1L is P △ 1H is P △ 2L is P △ TB s P Here, Z is Zero, P is Positive, PS is Positive Sma
ll and N are Negative and NS are Negative Small membership functions.

13 (a) to 13 (c) show membership functions (consequent part) of the inference rule using the result of charbed shape recognition. In each figure, the horizontal axis represents the normalized domain. , The vertical axis represents the degree. Inference Rule Using Floating Dryness Recognition Result The fuzzy controller 4 uses the output result of the neural network for recognizing the floating dryness as the antecedent part of the logical condition, and performs the fuzzy inference using the following inference rules to obtain Δ1H. , Δ2L, ΔTB are calculated. R ₂₁ : IF floating dry amount is zero THEN Δ1H is Z Δ2L is Z ΔTB is Z R ₂₂ : IF floating dry amount is small THEN Δ1H is NS Δ2L is NS ΔTB is NS R ₂₃ : IF floating dry Quantity is large THEN Δ1H is NB Δ2L is NB ΔTB is NB These membership functions are shown in FIG. Inference Rule Using SO ₂ Measured Value The fuzzy controller 2 determines the level of the SO ₂ measured value (the antecedent part), and calculates Δ1L and ΔTB by fuzzy reasoning using the following inference rules. R ₃₁ : IF SO ₂ is zero THEN Δ1L is Z ΔTB is Z R ₃₂ : IF SO ₂ is low THEN Δ1L is PS ΔTB is PS R ₃₃ : IF SO ₂ is high THEN Δ1L is PB ΔTB is PB These membership functions are shown in FIG. Inference Rule Using NOx Measurement Value The fuzzy controller 3 determines the level of the NOx measurement value (the antecedent part), and calculates Δ2H by fuzzy inference using the following inference rules. R ₄₁ : IF NOx is zero THEN Δ2H is Z R ₄₂ : IF NOx is low THEN Δ2H is NS R ₄₃ : IF NOx is high THEN Δ2H is NB FIG. 16 shows these membership functions.

According to this structure, the fuzzy inference unit 27 uses fuzzy logic to recognize the results of the charbed shape recognizing means 22 and the suspended matter amount recognizing means 25, and the SO ₂ concentration of the exhaust gas concentration detecting means 26 and the like. , NOx
We aim to improve the recovery rate of chemical raw materials while optimally and stably operating the recovery boiler from a comprehensive perspective by making appropriate judgments according to a number of recognition signals, such as concentration detection signals, and detection signals.
Further, there is an advantage that it is possible to suppress harmful components in the exhaust gas and prevent the deterioration of the environment.

[0033]

However, in order to actually install the recovery boiler optimum control device as described above in the plant and start the operation, the membership function of fuzzy inference, the scaling factor, the inference rule, etc. are set. It is indispensable to define concretely according to the actual operation and adjust so that the target process control is performed. Since there are many fuzzy parameters to be adjusted in such an assembly and adjustment work, it takes a long time for the fuzzy controller to perform the control operation that reflects the operation experience of the plant operator.

Therefore, it is an object of the present invention to provide a recovery boiler optimum control device capable of facilitating the adjustment of fuzzy parameters in the fuzzy inference part of the recovery boiler control device.

[0035]

In order to achieve the above object, the optimum control device for a recovery boiler according to the present invention injects black liquor discharged from the chip digestion process of a pulp manufacturing plant into the furnace to cause char to the bottom of the furnace. By combusting the char bed by supplying combustion air to a plurality of places in the furnace together with forming a bed, a recovery boiler that controls the combustion of a recovery boiler that generates steam and recovers a chemical raw material for chip digestion In the optimum control device, the shape of the above-mentioned char bed that has been classified in advance is stored, and to what extent the current char bed fits each classified char bed shape based on the char bed shape detection signal that represents the current char bed shape. And a charbed shape recognition means for generating a plurality of charbed shape conformity signals representing conformity for each shape. It is set in the fuzzy inference unit that calculates at least the black liquor temperature and the operation amount of the combustion air suitable for the operation of the recovery boiler using the fuzzy logic based on the charbed shape conformity signal and the fuzzy inference unit. Fuzzy parameter setting means for transferring the input fuzzy parameters to the fuzzy inference unit, charbed shape conformity storage means for storing the conformity of each of the charbed shape conformity signals over time, and the fuzzy inference Fuzzy inference result storage means for storing the fuzzy inference results calculated by the unit, fuzzy parameter storage means for storing the fuzzy parameters set in the fuzzy inference unit, and storage in the charbed shape conformance storage means Inference stored in the fuzzy inference result storage means Characterized by comprising an image display unit capable of displaying fuzzy parameters stored in the fuzzy parameter storage means together with results with time or fixedly displayed on the same screen a.

[0036]

When the contents of a plurality of fuzzy parameters and the values of the operating parameters of the recovery boiler such as the shape of the char bed and the values of the inference result of the fuzzy inference unit are memorized with time, and the temporal change of these operating parameters is displayed. In addition, by displaying the contents of the fuzzy parameters on the same screen, it becomes possible to visually understand the mutual relationship between the setting of the fuzzy parameters and the operating state of the recovery boiler, and the fuzzy reasoning section of the recovery boiler optimum control device It makes the selection of fuzzy parameters as easy as possible.

[0037]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS In the present invention, a screen for adjusting fuzzy parameters and a screen for displaying plant operation history data are provided so that the relationship between plant operating states and parameter set values can be visually grasped. , It makes it easy to transfer the operator's judgment and experience of plant operation to the fuzzy controller.

An embodiment of the present invention will be described below with reference to FIG. In the figure, parts corresponding to those in FIG. 7 are designated by the same reference numerals, and description of such parts will be omitted.

In FIG. 1, fuzzy parameters for inputting fuzzy parameters such as membership functions, scaling factors, inference rules, etc. for fuzzy inference to the fuzzy controllers 1 to 4 of the fuzzy inference unit 27 are input. A setting unit 51 is provided. The fuzzy parameter setting unit 51 is configured as, for example, an operation control panel, and the parameter setting state can be visually recognized by a scale on the panel or a display unit, and various parameters can be output by a keyboard switch, a dial type knob, or the like. Has been done.

The output of the fuzzy parameter setting unit 51 is supplied to the image display control unit 52, and the parameter setting state is stored in the fuzzy parameter storage area of the large-capacity memory device 53 at predetermined sample time intervals. The output of the fuzzy parameter setting unit 51 is the image display control unit 52.
It is supplied to the corresponding fuzzy computer of the fuzzy inference unit 27 via, and the parameter value is set.

The image display control section 52 is supplied from the charbed shape recognizing means 22 with shape recognition signals TYPE0 to TYPE6 representing the charbed shape which is an operating parameter. These levels (goodness of fit) are sampled at predetermined sampling time intervals and stored in the charbed shape storage area of the memory device 53.

Further, the floating amount recognizing means 25 supplies a signal indicating "zero", "small", and "large" of the floating amount, which is an operating parameter, to the image display control section 52, and their levels (degree of conformity). Are sampled at a predetermined sampling time interval and stored in the floating amount storage area of the memory device 53.

Further, from the exhaust gas concentration detecting means 26, two exhaust gas concentration detection signals representing the NOx and SO ₂ concentrations in the exhaust gas, which are operating parameters, are supplied to the image display control section 52, and their levels are supplied. Are sampled at predetermined sampling time intervals and stored separately in the NOx storage area and the SO ₂ storage area of the memory device 53.

Further, a signal Δ indicating the change of the primary low air flow rate which is the fuzzy control output from the fuzzy inference unit 27.
1L, a signal representing the change in the primary high air flow rate Δ1H, 2
The signal 2L indicating the change amount of the next low air flow rate, the signal 2H indicating the change amount of the secondary high air flow rate, and the signal ΔTB indicating the change amount of the black liquor temperature are supplied to the image display control unit 52, and sampling is performed at a predetermined sampling time interval. The fuzzy manipulated variable storage areas Δ1L, Δ1H, 2 for each signal of the memory device 53 are stored.
It is stored in L, 2H and ΔTB.

Here, the fuzzy parameter storage area, the charbed storage area, the floating amount storage area, the NOx storage area, and the SO ₂ storage area formed in the memory device 53 are fuzzy parameter storage means and charbed shape conformance storage, respectively. Means, means for storing suspended solids, means for storing exhaust gas concentration. The fuzzy operation amount storage area corresponds to fuzzy inference result storage means.

The image display control unit 52, for example, processes the data to be displayed as an image stored in each storage area of the memory device 53, converts it into two-dimensional screen image information, and controls the image display system. Microprocessor that performs conversion, stores converted two-dimensional information as pixels, and CR
Bit map display memory, such as T, supplied to the image display 54, a control program, a ROM storing a plurality of operating parameters and fuzzy parameters and fuzzy control outputs related to each operating parameter, and a designated character / symbol pattern Character ROM generated and supplied to the above display memory, light pen logic interface for inputting commands from the display screen, movement of cursor on display screen and input of commands corresponding to display contents of cursor position on screen It is composed of a mouse interface and the like for performing. Image display control unit 52 and image display 5
Reference numeral 4 corresponds to an image display means.

The fuzzy parameter setting section 51, the image display control section 52, the memory device 53 and the image display 54.
The fuzzy parameter adjustment support unit 50 is thus formed. Other configurations are similar to those of the conventional device.

Next, the operation of the above fuzzy parameter adjustment support unit 50 will be described with reference to the drawings.

When the operator of the recovery boiler selects the charbed shape as the operation parameter from the keyboard of the fuzzy parameter setting section 51, the microprocessor refers to the file stored in the ROM in advance using the charbed shape as an index, and the charbed Determine fuzzy parameters and fuzzy control outputs related to the shape.
For example, fuzzy parameters set in the fuzzy controller 1 for controlling the charbed into an ideal shape, and fuzzy control outputs Δ1L, Δ1H, Δ2L, ΔTB.
Know that is relevant. The microprocessor reads the accumulated data relating to these operating parameters and fuzzy control outputs from the memory 53 up to a certain time range before the present, and forms a graph showing the values of the parameters over time in a predetermined graph display area on the display memory. To do. Also, the contents of the currently set fuzzy parameters are written in a predetermined table display area on the display memory. When it is instructed to display the contents of the fuzzy parameters at a predetermined time earlier than the present time, the microprocessor reads the data at this time from the corresponding storage area of the memory 53 and stores it in the display memory. Write in the specified table display area. The contents of this display memory are transferred to the image display 54 and displayed on the screen.

FIG. 2 shows the image display 54 when the charbed shape is selected as the operation parameter of the recovery boiler.
The example of the screen display of is shown.

In the upper left part of the screen of the image display 54, the fitness of each of the shape recognition signals TYPE0 to TYPE6, that is, how close the shape of the charbed, which is an operating parameter, is to the classified shape, corresponds to the plant operating time. And is displayed as a curve over time. For example, in FIG. 2, the shape of the char bed is TYPE2 (shown in FIG. 9B),
Then, the shape of TYPE3 (shown in FIG. 9C) is found to be effective.

These shape recognition signals TYPE0-TYPEP
E6 is multiplied by the input gain set for each signal in the input adjustment section of the fuzzy controller 1 and sent to the fuzzy inference operation section.

In the selection state of the membership function in the fuzzy inference operation section of the fuzzy controller 1, the shape recognition signals TYPE0 to TYPE6 are set in the upper right part of the display screen, and the air change amounts Δ1L, Δ1H, Δ2L, Δ2H and It is displayed as a table on a matrix with the black liquor temperature adjustment amount ΔTB as the vertical axis.

The operator of the recovery boiler selects the positions of the membership functions arranged in a matrix with the light pen or the mouse on the display screen, and the fuzzy parameter setting section 51 selects the membership function types Z, P and P.
By inputting any of S, N and NS, it is possible to change the membership function that has already been selected.

At the lower right of the screen of the image display 54, the output gain values a ₁ , a ₂ , a ₃ and a ₄ set in the output adjusting section of the fuzzy controller 1 are signals Δ indicating the change in the air flow rate. 1L, △ 1H, △ 2L and black liquor temperature change △ TB
Corresponding to is shown. The value of each output gain is input to the output adjustment circuit of the fuzzy controller 1 by inputting the position on the screen with a light pen or mouse on the screen and inputting the value from the keyboard of the fuzzy parameter setting unit 51. It is set as a value. In the input adjustment unit of the fuzzy controller 1, when the same input gain is set for the shape recognition signals TYPE0 to TYPE6, the display of the input / output gain can be omitted.

At the lower left of the screen of the image display 54, the primary low air flow rate change amount Δ1L, the primary high air flow rate change amount Δ1H,
Each value of the secondary high air flow rate change amount Δ2L is shown by a curve over time corresponding to the plant operation. Black liquor temperature setting change amount △
TB is also shown as a curve over time. The secondary high air flow rate change amount Δ2H is excluded from the control target because it has little influence on the shape of the char bed.

When the operator selects the floating dry amount as the operation parameter from the keyboard of the fuzzy parameter setting section 51, the image display control section 52 performs the same data processing as the processing for displaying FIG. A fuzzy inference parameter adjustment screen using the floating dry amount recognition result as shown in FIG. 3 is formed in a display memory and supplied to the image display 54.

FIG. 3 shows an example of the display screen of the image display 54 in this case, and the suitability for the floating dry amount recognition results “zero”, “small”, and “large” is shown in the upper left of the screen. The change is shown as a curve. The membership function selected in the fuzzy inference operation unit of the fuzzy controller 4 is displayed on the upper right of the screen. The type of membership function is changed by designating a position on the screen with a light pen or the like and inputting the type of function from the fuzzy parameter setting unit 51.

At the bottom right of the screen, the fuzzy controller 4
The gain values a _{5 to} a ₇ set in the output adjustment circuit of are respectively the primary high air flow rate change amount Δ1H and the secondary low air flow rate 2
L and the black liquor temperature change amount ΔTB are set correspondingly.
These output gains can also be changed by inputting a position on the screen with a light pen or the like and inputting values from the keyboard of the fuzzy parameter setting unit 51.

At the lower left of the screen, the primary high air flow rate change amount Δ
1H, secondary low air flow rate change amount △ 2L, black liquor temperature change amount △
The change in TB over time is shown by the curve.

From the keyboard of the fuzzy parameter setting section 51, the operator can select S in the exhaust gas as an operating parameter.
When the density of O ₂ is selected, the image display control unit 52 performs the same data processing as the processing for displaying FIG. 2 described above, and the screen as shown in FIG. 4 is displayed.

At the upper left of this screen, the concentration of SO _{2 in} the exhaust gas is shown as a curve over time. At the upper right of the screen, the types of membership functions set in the fuzzy inference unit are shown. This type can be changed by inputting the position on the screen with a light pen or the like and designating the type of membership function from the fuzzy parameter setting section 51.

At the bottom right of the screen, the fuzzy controller 2
Gain for each signal set to the input adjustment unit and an output adjustment of the values b _8, b _9, and a _8, a ₉ is shown.

At the lower left of the screen, the primary low air flow rate change amount Δ
The change over time in 1 L and the black liquor temperature change amount ΔTB is shown by a curve. The signals other than the signals Δ1L and ΔTB are not shown because they are less related to the increase / decrease in SO ₂ .

From the keyboard of the fuzzy parameter setting section 51, the operator can set the N in the exhaust gas as an operating parameter.
When the density of Ox is selected, the image display control unit 52 performs the same data processing as the processing for displaying FIG. 2 described above, and the screen as shown in FIG. 5 is displayed. At the upper left of this screen, the concentration of NOx in the exhaust gas is shown as a curve over time.

On the upper right of the screen, the fuzzy controller 3
The kind of membership function set in the fuzzy inference part of is shown. When the influence of the change amount other than the secondary high air flow rate change amount Δ2H on the reduction of NOx is small, only Δ2H may be displayed as shown in the figure.

At the lower right of the screen, the input gain b ₁₀ and the output gain a ₁₀ set in the input adjustment unit and the output adjustment unit of the fuzzy controller 3 are displayed. These gain values can be changed by inputting a position on the screen with a light pen or the like, and by inputting a numerical value from the keyboard of the fuzzy parameter setting unit 51.

At the lower left of the screen, the secondary high air flow rate change amount Δ2
The change in H with time is shown by a curve.

As described above, the fuzzy parameters such as the kind of membership function set in the fuzzy controller and the input / output gain are displayed on the screen of the image display 54 of the fuzzy parameter adjustment support unit 50, and at the same time, the fuzzy controller outputs. And the controlled amount of the recovery boiler are displayed over time on the same screen. In addition, since the setting state of the fuzzy parameters at various times of various operating parameter curves displayed over time can be read and confirmed from the memory device 53, quantitative analysis and confirmation of the influence of the parameter setting on the operating parameters can be performed. Will also be easier. Since the selection state of the fuzzy parameters is stored in the memory over time, it is possible to display this selection state as a graph over time on the display screen.

Therefore, the operator of the recovery boiler can easily determine the fuzzy parameters while confirming the operation state of the recovery boiler to realize the optimum operation control of the recovery boiler by the fuzzy control, and install the recovery boiler plant. Full-scale operation of the subsequent plant can be accelerated as soon as possible.

The fuzzy parameter adjustment support unit 50
Can be removed after the fuzzy parameter setting is completed, and this can be used again and again for the fuzzy parameter adjustment when installing another recovery boiler.

Further, it is possible to provide a plurality of image display devices 54 and simultaneously display the adjustment screens associated with each fuzzy controller. In such a case, the plant operator can overlook the entire control flow rate and the like, and it is easier to realize the operation control of the recovery boiler that has a high recovery efficiency and suppresses the exhaust gas concentration and is generally preferable.

It is also possible to provide a touch panel sensor on the screen of the image display 54 and display a picture of a keyboard on the screen so that all fuzzy parameter settings are instructed from the screen.

[0074]

As described above, in the optimum control system for the recovery boiler of the present invention, at least historical data of the operation parameters of the recovery boiler and the fuzzy parameters are displayed on the same screen over time, and the fuzzy parameters are set. Since an adjustment screen is provided to visually grasp changes in operating parameters with time, the recovery boiler operator can use fuzzy parameters such as membership functions of fuzzy inference, scaling factors, and inference rules. Can be adjusted more easily. Further, it is possible to shorten the plant assembly adjustment time as much as possible.

[Brief description of drawings]

FIG. 1 is a functional block diagram showing an embodiment of the present invention.

FIG. 2 is a diagram showing a first display example of an image display device 54.

FIG. 3 is a diagram showing a second display example of the image display device 54.

FIG. 4 is a diagram showing a third display example of the image display device 54.

FIG. 5 is a diagram showing a fourth display example of the image display device 54.

FIG. 6 is a configuration diagram showing a configuration of a recovery boiler.

FIG. 7 is a functional block diagram showing a configuration of a conventional control device.

FIG. 8 is a functional block diagram showing the configuration of a charbed shape recognition unit 22.

FIG. 9 is a diagram in which the shapes of char beds are classified.

FIG. 10 is a functional block diagram showing a configuration of a suspended matter amount recognition unit 25.

FIG. 11: Classification of floating dry amount.

FIG. 12 is a functional block diagram showing the configuration of a set value changing unit 27.

FIG. 13 is a diagram showing an example of a membership function used for fuzzy inference based on a result of recognition of a shape of a charbed.

FIG. 14 is a diagram showing an example of a membership function used for fuzzy inference based on a floating dry amount recognition result.

FIG. 15 is a diagram showing an example of a membership function used for fuzzy inference based on SO ₂ measurement values.

FIG. 16 is a diagram showing an example of a membership function used for fuzzy inference based on NOx measurement values.

[Explanation of symbols]

 1 Black Liquor Heater 2 Black Liquor Temperature Control Valve 4 Recovery Boiler 5 Char Bed 22 Char Bed Shape Recognition Means 25 Floating Amount Recognition Means 26 Exhaust Gas Concentration Detection Means 27 Fuzzy Inference Part 28 Air Flow Rate Setting Part 30 Set Value Change Part 50 Fuzzy Parameter Coordination Support Department

Claims (3)

[Claims] 1. The char bed by injecting black liquor discharged from a chip cooking process of a pulp manufacturing plant into a furnace to form a char bed at the bottom of the furnace and supplying combustion air to a plurality of places in the furnace. Is an optimal control device of a recovery boiler that controls the combustion of a recovery boiler that recovers the chemical raw material for chip digestion by producing steam, and stores the pre-classified shape of the char bed. Of multiple charbed shapes that represent the goodness of fit for each shape by determining how well the current charbed fits each classified charbed shape based on the charbed shape detection signal A charbed shape recognition means for generating a degree signal, and based on at least the charbed shape matching degree signal,
A fuzzy inference unit that calculates at least the temperature of the black liquor and an operation amount of the combustion air suitable for operating a recovery boiler using a fuzzy logic; and a fuzzy parameter input to be set in the fuzzy inference unit. Fuzzy parameter setting means for transferring to the fuzzy inference section, charbed shape conformity storage means for storing the conformity of each of the charbed shape conformity signals with time, and the inference result calculated by the fuzzy inference section for time Fuzzy inference result storage means for storing the fuzzy inference result, fuzzy parameter storage means for storing the fuzzy parameters set in the fuzzy inference unit, each signal value stored in the charbed shape conformance storage means, and the fuzzy inference result When the inference result stored in the storage means is displayed on the same screen over time or fixedly , The optimal control system of the recovery boiler, characterized in that an image display unit capable of displaying fuzzy parameters stored in the fuzzy parameter storage means. 2. The char bed by injecting black liquor discharged from a chip cooking process of a pulp manufacturing plant into a furnace to form a char bed at the bottom of the furnace and supplying combustion air to a plurality of places in the furnace. Is an optimal control device of a recovery boiler that controls the combustion of a recovery boiler that recovers the chemical raw material for chip digestion by producing steam, and stores the pre-classified shape of the char bed. Of multiple charbed shapes that represent the goodness of fit for each shape by determining how well the current charbed fits each classified charbed shape based on the charbed shape detection signal The shape of the char bed that generates the degree signal, and the amount of suspended matter in the upper part of the furnace that has been classified in advance are stored to detect the amount of suspended matter that is present. Floating quantity recognition means for determining how much the current floating quantity conforms to each classification based on the signal and generating a plurality of floating quantity conformity signals indicating the conformity for each classification; and at least the charbed shape A fuzzy inference unit for calculating at least the temperature of the black liquor and the manipulated variable of the combustion air suitable for the operation of the recovery boiler by using different fuzzy logics based on the fitness signal and the floating material fitness signal, Fuzzy parameter setting means for transferring a fuzzy parameter input to be set in the fuzzy inference unit to the fuzzy inference unit, and a charbed shape adaptability for storing the adaptability of each of the charbed shape adaptability signals over time. Storage means, floating material quantity storage means for storing the fitness of each of the floating material fitness signals over time, and the estimation calculated by the fuzzy inference unit. At least one of fuzzy inference result storage means for storing results over time, fuzzy parameter storage means for storing fuzzy parameters set in the fuzzy inference unit, at least one of the charbed shape conformance storage means and the suspended matter amount storage means Each of the signal values stored in one of the storage means and the fuzzy inference result stored in the fuzzy inference result storage means are displayed on the same screen over time or in a fixed manner, and the fuzzy parameter storage means is stored. An optimal control device for a recovery boiler, comprising: an image display means capable of displaying parameters. 3. The char bed by injecting black liquor discharged from a chip cooking process of a pulp manufacturing plant into a furnace to form a char bed at the bottom of the furnace and supplying combustion air to a plurality of places in the furnace. Is an optimal control device of a recovery boiler that controls the combustion of a recovery boiler that recovers the chemical raw material for chip digestion by producing steam, and stores the pre-classified shape of the char bed. Of multiple charbed shapes that represent the goodness of fit for each shape by determining how well the current charbed fits each classified charbed shape based on the charbed shape detection signal Exhaust gas concentration indicating the concentration of exhaust gas components discharged from the furnace Exhaust gas concentration detection means for generating a detection signal, and based on at least the charbed shape conformity signal and the exhaust gas concentration detection signal, at least the black liquor of at least the optimum for the operation of the recovery boiler using different fuzzy logic A fuzzy inference unit that calculates a temperature and the manipulated variable of the combustion air; a fuzzy parameter setting unit that transfers a fuzzy parameter input to be set in the fuzzy inference unit to the fuzzy inference unit; Degree shape fitness degree storing means for storing the fitness degree of each degree signal over time, exhaust gas concentration storing means for storing the exhaust gas concentration detection signal over time, and fuzzy inference calculated by the fuzzy inference section Fuzzy inference result storage means for storing results over time, and a fuzzy parser set in the fuzzy inference unit. A fuzzy parameter storage means for storing parameters, each signal value stored in at least one of the charbed shape conformity storage means and the exhaust gas concentration storage means, and the fuzzy inference result storage means. Optimum of the recovery boiler, characterized in that the fuzzy inference result and the fuzzy parameter stored in the fuzzy parameter storage means are displayed on the same screen with time or fixedly, and image display means capable of displaying the fuzzy parameter. Control device.