JPH06137107A - Turbine control device - Google Patents

Abstract

PURPOSE:To dispense with fine adjustment according to characteristic or secular change of a turbine and impart high controllability. CONSTITUTION:This turbine control device 1 is constituted of a speed command preparation part 13 to search a command speed at present and an estimated command speed after a short time out of a target speed TS and a target acceleration TA, a speed estimation part 12 to search an estimated speed after a short time out of an actual speed AS at present, adder-subtractor devices 14, 15 to search the present value e (0) and the estimated value e (1) of control deviation, a fuzzy inference part 16 to output the variation of the position of a steam valve from the e (0) and e (1), and a valve command preparation part 17 to prepare the position command of the steam valve according to variation of the position of the steam valve. Thereby, high controllability is obtained so as to enable turbine control of high accuracy, even without finely adjusting the control device against secular change of the turbine or change of the operating condition.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a turbine control device for controlling steam turbines and gas turbines of nuclear and thermal power plants, etc., and in particular, fine adjustment of the control device according to the characteristics and aging of the turbine is unnecessary. The present invention relates to a turbine control device suitable for smoothing and improving responsiveness and controllability.

[0002]

2. Description of the Related Art FIG. 2 shows an example of a conventional turbine control device. At the time of turbine acceleration, the target acceleration TA
And the deviation from the actual acceleration obtained from the actual speed AS by the differentiator 31 are input to the speed PI calculator (PIV) 32, and the steam valve of the turbine 2 is operated by the speed PI calculator and integrator 35. Is generated and the valve is opened / closed by the command signal VP to adjust the steam flow rate, whereby the actual acceleration is controlled to be a value close to the target acceleration TA. The speed increases and the target speed TS and the actual speed A
When the deviation of S decreases, the switching logic 34 switches to the speed PI calculator 33, and constant speed control is performed so as to eliminate the deviation between the target speed TS and the actual speed AS.

In the above conventional turbine control device,
It was necessary to finely adjust the set values of the PI computing unit, the control loop parameter, and the switching logic according to the characteristics (capacity, inertia) of the turbine body, aging, and the like.

As a control method for improving such a problem, there is a fuzzy control which has recently received attention. Fuzzy control has been put into practical use since the 1980s, and is currently applied to tunnel ventilation control of motorways, train fixed position stop control, control of home appliances, and the like. In fuzzy control, a control operation is described by a control rule including ambiguity such as "large" or "small", and a control target is controlled by determining a control command using fuzzy reasoning. Due to the ambiguity of the control rules, automatic control that can flexibly respond to differences in the characteristics of the controlled object and changes over time becomes possible.

As one of conventional examples of fuzzy control, the publication "Systems and Control," Vol. 28, No. 7, 442 to 4 is used.
As described on page 46, there is a method of using a fuzzy inference based on a deviation between a measured value of a controlled variable and a target value (hereinafter referred to as a control deviation) and a time change rate of the control deviation.

As another conventional example, pages 1973 to 880 of Vol. 19 No. 11 of the Society of Instrument and Control Engineers,
Predictive fuzzy control has been proposed. This is a method of evaluating a plurality of control rules including predictions for control execution results in addition to ambiguity, and selecting the control rule with the highest evaluation value. A better response than the conventional fuzzy control example. And controllability.

Further, the gas turbine control device described in Japanese Patent Laid-Open No. 3-258924 has a gasifier reed mode in which the load is controlled by the gasifier fuel input amount and the system pressure is controlled by the fuel valve opening degree. On the contrary, the operation amount of two control modes, gas turbine reed mode in which the system pressure is controlled by the gasifier fuel injection amount and the load is controlled by the fuel valve opening, is proportionally divided by a weighted average to generate a new operation amount. At this time, there is provided means for obtaining the weight of the weighted average by fuzzy inference based on two control deviations, a load deviation and a system pressure deviation. As a result, it becomes possible to perform automatic control that can precisely and flexibly respond to changes in the operating mode of the plant and aging of the turbine.

[0008]

In each of the conventional fuzzy controls described above, it is necessary to increase the number of control rules in order to improve the responsiveness and controllability of the controlled variable. Therefore, there is a problem that the calculation time required for control increases, and it is difficult to sufficiently improve responsiveness and controllability. It was mentioned above that the predictive fuzzy control has good responsiveness and controllability.
However, in this control method, it is necessary to obtain evaluation values for all control rules and foresee the execution result of each control, which complicates the algorithm. For this reason,
If the number of control rules is increased in order to improve the responsiveness of the control amount, the calculation time will be longer than that of other conventional fuzzy control, and the good responsiveness and controllability inherent in the predictive fuzzy control will be offset. . Also, if you do not choose the evaluation rule well,
There is also a problem that desired responsiveness and controllability cannot be obtained.

The present invention has been made in view of the above problems, and eliminates the need for fine adjustment of the control device according to the characteristics and aging of the turbine, and has high responsiveness and controllability. The purpose is to provide a device.

[0010]

Means for Solving the Problems The above object is to provide a turbine control device with a command value at a present time of a control amount such as a rotation speed and acceleration, an amount of power generation according to a load, and a command value at a future time point after the present time. It is achieved by providing means for controlling the turbine by fuzzy inference, based on the predicted value, the current value, and the predicted value at a future time.

[0011]

If the above means is applied, the fuzzy inference unit can be input to
It includes information about transient influences such as disturbances and dead time on the control output at future time points after the present time, and sets the control rule according to the transient influences so that disturbances and dead time are predicted. Fuzzy control that can cope with the influence of Therefore, since it is not necessary to increase the control rules or use complicated algorithms in the fuzzy inference unit, the calculation time required for control does not increase unlike the conventional fuzzy control. Therefore, it is possible to obtain a turbine control device having high responsiveness and controllability and capable of automatic control capable of precisely and flexibly accommodating the operating form of the plant and the secular change of the turbine.

[0012]

Embodiments of the present invention will be described below with reference to the drawings. The same symbols are attached to the same components and equivalent components in the drawings.

FIG. 1 is a block diagram showing a steam turbine controller according to an embodiment of the present invention. The turbine control device 1 inputs a target speed TS, a target acceleration TA, and an actual speed AS at the present time and outputs a change amount of a position of a steam valve of the turbine 2 by fuzzy inference, and a fuzzy control device portion 11 of the steam valve. It comprises a valve command creating unit 17 which creates a position command for the steam valve in accordance with the amount of change in position. The fuzzy control unit 11 calculates a command speed at present and a predicted command speed after a short time from a target speed TS and a target acceleration TA, and a speed command creation unit 13 calculates a predicted speed after a short time from an actual speed AS. Prediction unit 12, deviation e between current command speed and actual speed

An adder / subtractor 14 for obtaining [0], an adder / subtractor 15 for obtaining a deviation e [1] between the predicted command speed and the predicted speed after a short time, and the deviation e

The fuzzy inference unit 16 outputs a change amount of the position of the steam valve by fuzzy inference from [0] and e [1]. The fuzzy inference unit 1
6 is, for example, “deviation e

[0] is close to 0 and the deviation e
If [1] is negative and large, the amount of change in the position of the steam valve is negatively increased (that is, the steam valve is greatly closed), and a control rule storage unit 161 that stores a plurality of control rules, and , Deviation e

[0], the deviation e [1] and the value of the amount of change in the position of the steam valve correspond to "close to 0", "negative and large", etc. included in the control rule in the interval [0,
1] has a membership function storage unit 162 that stores a plurality of membership functions represented by real values. Next, a control method in the control apparatus of the present embodiment will be described by taking turbine speed increasing control and constant speed control as examples.

With the target speed TS and the target acceleration TA set by the operator as inputs, the speed command preparation unit 13 obtains the command speed at the present time and the predicted command speed after a short time, and the target acceleration TA is aimed at the target speed TS. A command speed pattern is created as a pattern for speeding up. Further, the actual speed AS of the turbine 2 measured at the present time is input, and the predicted speed after a short time is obtained by the speed prediction unit 12. Furthermore, the adder / subtractors 14 and 15 are used to deviate the deviation e between the current command speed and the actual speed e.

[0] (= command speed-actual speed), deviation e between predicted command speed and predicted speed after a short time
[1] (= predicted command speed-predicted speed) is created. Here, the predicted command speed after a short time is obtained as the current target speed and the target acceleration continue to be selected, and the predicted speed after a short time is the speed change from the speed before several sampling and the actual speed at the present time. Obtained as being linear. Deviation e above

The fuzzy inference unit 16 obtains the amount of change in the position of the steam valve by inputting [0] and e [1]. The turbine speed is controlled by operating the steam valve and adjusting the steam flow rate in accordance with the turbine steam valve position command created by the valve command creation unit 17 based on the amount of change in the steam valve position.

Here, the operation of the fuzzy inference unit 16 in this embodiment will be described.

The control rule stored in the control rule storage unit 161 is "deviation e.

If [0] is A1 and the deviation e [1] is A2, the change amount of the position of the steam valve should be B. ”A1, A2 and B in the control rule are It is an ambiguous variable (hereinafter referred to as fuzzy variable) that expresses the magnitude of change in position and its magnitude in language. In this embodiment, five types of fuzzy variables, "positive and large", "positive and medium", "close to zero", "negative and medium", and "negative and large", are used respectively and PB (PositiveBig) Abbreviation), PM (abbreviation of Positive Medium), Z (abbreviation of Zero), NM (abbreviation of Negative Medium) and NB (Negative)
Short for Big).

FIG. 3 shows membership functions stored in the membership function storage unit 162 of the fuzzy inference unit 16 and corresponding to the fuzzy variables. Horizontal axis shows deviation e

[0] and e [1] and the amount of change in the position of the steam valve are shown and divided into positive and negative regions with 0 as the center. However, since the value on the horizontal axis differs depending on the turbine, the description of specific values is omitted. The vertical axis represents the degree to which each deviation and the amount of change in the position of the steam valve correspond to each fuzzy variable (denoted as grade in the figure), and is represented by a numerical value from 0 to 1.

FIG. 4 shows a control rule. In this embodiment,
Nine control rules are used. For example, rule number 1
Is the deviation e

If [0] is Z and the deviation e [1] is NB, the change amount of the position of the steam valve should be NB. " In other words, if the turbine speed is predicted to be close to the command speed at the present time and to be considerably higher than the command speed in a short time, the steam valve should be closed greatly.

In the fuzzy reasoning section 16, first, the deviation e

By inputting [0] and e [1], the degree (grade) where each deviation corresponds to the above-mentioned five types of fuzzy variables is obtained from the membership function of FIG. The grade of each of these deviations,
From the control rules shown in FIG. 4 and the membership function with respect to the variation of the steam flow rate, the variation of the position of the steam valve is output using a known fuzzy inference method. As a fuzzy inference method, there is the Min-Max centroid method, which is well known as the method of Professor Mamdani, described in Proceeding of IEEE Vol. 121, pp. 1585 to 1588.

The above is the operation of the fuzzy inference unit 16, but the membership function, the control rule, and the fuzzy inference method are not limited to the above.

Next, an operation simulation of turbine control by the control device of the above embodiment performed by the inventors will be described.

FIG. 5 is a model of the turbine used in the operation simulation of this embodiment. It is a model consisting of a combination of the elements shown in each block diagram, which inputs the position command of the steam valve and outputs the actual position and the actual speed of the steam valve.
In the figure, K1, H1, L1, B, H2, T2, K
2, K3 and T3 are parameters.

FIGS. 6, 7 and 8 show simulation results. The values of the above parameters are also shown. Figure 6
On the other hand, the parameters H2 and T3 are changed in FIG. 7, and the parameters L1, H2 and T3 are changed in FIG. H2 and L1 relate to the characteristics of the steam valve (for example, the relationship between the valve position and the steam flow rate), and T3 represents the time constant of the turbine. Therefore, changes in these parameters correspond to changes in the characteristics of the turbine and valves due to changes over time. In the figure, 1 is the output of the fuzzy inference unit (that is, the amount of change in the position of the steam valve), 2 is the position command of the steam valve, 3 is the actual position of the steam valve, 4 is the speed command, and 5 is the actual speed. In each simulation result, the actual speed almost matches the command speed. That is, it can be seen that the control device of the present embodiment exhibits high controllability even if the characteristics of the turbine or the valve change due to aging or the like.

As described above, the fuzzy inference in this embodiment is executed based on the present value of the control deviation of the turbine speed and the predicted value after a short time, so that the number of control rules is not increased and the algorithm is not complicated. However, high controllability is obtained. Further, due to the flexibility of fuzzy control, stable and highly accurate speed control is possible without fine adjustment of the control device when the characteristics or operating conditions of the turbine change. Also, since the rate of change in control deviation over time is not used,
It also has the effect of being less susceptible to noise.

The turbine control device of the present invention can be applied not only to speed control but also to load control. in this case,
The power generation amount command from the outside and the output of the generator are input to the turbine control device. Further, the fuzzy control unit of the above embodiment is effective not only for turbines, but also for control targets including dead time and control targets that require fine adjustment of the control device due to secular change, etc. High controllability can be obtained. Furthermore, it is possible to use the fuzzy control method in the turbine control device of the present invention and the conventional non-fuzzy control method together, and to switch between the two as appropriate.

[0026]

As described in detail above, according to the present invention,
When the turbine has changed over time or the operating conditions have been changed, high controllability can be obtained and high-precision turbine control is possible without fine adjustment of the control device. Further, the resistance to disturbance of the power plant is improved and the reliability of the plant is improved. Further, fine adjustment of the control device due to maintenance and inspection of the turbine is unnecessary, and the time required to restart the turbine can be shortened.

[Brief description of drawings]

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

FIG. 2 is a block diagram of a conventional turbine control device.

FIG. 3 is a membership function according to an embodiment.

FIG. 4 is a control rule according to the embodiment.

FIG. 5 is a turbine model used in the simulation.

FIG. 6 is a result of operation simulation of the example.

FIG. 7 is a result of operation simulation of the example.

FIG. 8 is a result of operation simulation of the example.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Turbine control device, 2 ... Turbine, 11 ... Fuzzy control device part, 12 ... Speed prediction part, 13 ... Speed command preparation part, 14 ... Adder / subtractor, 15 ... Adder / subtractor, 16 ... Fuzzy inference part, 17 ... Valve command Creation unit, 161 ... Control rule storage unit, 162 ... Membership function storage unit.

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification number Internal reference number FI technical display location G05B 13/04 9131-3H (72) Inventor Takumi Kawai 5-2-1 Omika-cho, Hitachi-shi, Ibaraki No. Incorporated company Hitachi Ltd. Omika factory (72) Inventor Sadao Yanagida 5-2-1 Omika-cho, Hitachi City, Ibaraki Prefecture Incorporated company Hitachi Ltd. Omika factory

Claims (2)

[Claims] 1. A turbine control device comprising: a means for creating a command value of the control quantity at the present time based on at least a target value of the control quantity; and a means for creating a predicted value of the command value at a future time point after the present time. , A means for obtaining a predicted value of the controlled variable at the future time point based on at least the current value of the controlled variable, a command value at the present time point of the controlled variable value, a predicted value of the command value at a future time point after the present time point, and a present value And a means for controlling the turbine by fuzzy inference based on a predicted value at a future time point. 2. The turbine controller according to claim 1, wherein the controlled variable is the speed of the turbine.