JPH09152903A - Method and device for generating plant operation paln - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make the time required for starting, stopping, etc., shortest while meeting operation restricting conditions, and environmental prescribed values by indicating the correction of the operation restricting conditions to a 2nd predicting means and optimizing them. SOLUTION: An optimizing means C is inferior in the speed and precision of prediction for a plant state corresponding to operation setting to a 1st predicting means A, but automatically searches for the optimum operation setting contents of the operation of the plant as to at least events at least one of start time, stop time, load variation time, and accident time by using a faster dynamic characteristic simulator and a 2nd predicting means B equipped with a function that can evaluate the operation restricting conditions faster and easier. Then an optimum operation adjusting means D compares the respective conditions and results of the 1st predicting means A and 2nd predicting means B and adjusts the evaluating means and decision reference value of operation restriction conditions in the 2nd predicting means B and further input conditions of the simulator and the input value of a model so that the differences between both of them become smaller.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a power plant such as a thermal power plant for starting, stopping, changing loads, and an operation control system at the time of an accident, and is generated in a steam generator such as a boiler, a steam turbine, and a gas turbine. Plant start, stop, load change, accident treatment, etc. in the shortest time while clearing all equipment usage limit values such as thermal stress and environmental regulation values such as plant emission NOx (hereinafter referred to as operation restriction conditions) The present invention relates to a plant operation plan creation device that enables automatic creation of an enabled optimum operation schedule. Especially,
It is suitable for various combined cycle power plants that combine a gas turbine, an exhaust heat recovery boiler, and a steam turbine.

[0002]

2. Description of the Related Art A conventional method will be described below by taking a process of starting a thermal power plant.

Conventionally, depending on the stop time before starting and the temperature condition of the equipment, the initial injection fuel amount into the boiler, the time function of temperature rise and pressure increase of main steam, the time function of turbine speed up and load increase are A method has been adopted in which a start-up schedule is determined and this start-up schedule is executed by a device-based or system-based control system provided in each system of the plant.

This most typical method is Electrical Wor
ld, Vol.165, No.6 paper "Thermal Stress Influenc
e Starting, Loading of Boilers and Turbines ". This method uniquely determines the starting schedule according to the initial conditions of a limited part of the plant. That is, boiler steam pressure, boiler outlet steam According to the temperature and the initial value of the steam turbine casing temperature, a method for determining the steam turbine acceleration rate, initial load, speed maintenance, and steam turbine warm-up time and load change rate due to load retention. Since it is not possible to predict the temperature rise characteristics of the boiler-generated steam, which is important in managing the thermal stress of the steam turbine, which is an operation restriction factor, before the start-up, the uncertainty is absorbed by giving a margin to the start-up schedule. Therefore, the startup schedule created tended to be longer than necessary.
The same applies to combined cycle power plants that combine a gas turbine and a steam turbine. In this case, the input energy to the plant is specified by determining the fuel input amount to the gas turbine, that is, the gas turbine acceleration rate and the load increase rate, instead of the fuel amount input to the boiler.

As another conventional method, US Pat.
The ones disclosed in 446,224 and U.S. Pat. No. 4,228,359 are known. These are intended to rapidly start the steam turbine while monitoring the thermal stress generated in the steam turbine in online real-time, but like the above-mentioned conventional method, there is no mention of the method for starting the boiler.

As a conventional method for reducing the start-up time of a boiler, the one described in Japanese Patent Laid-Open No. 59-157402 is known. This method aims to rapidly raise the temperature of steam generated by the boiler while monitoring the thermal stress generated in the boiler in real time online. However, this method makes no reference to the start of the steam turbine, and does not make any reference to the management or control of NOx emitted from the boiler. Therefore, there is no guarantee that the NOx emission amount due to the rapid temperature rise of boiler-generated steam can be limited to the environmental regulation value level or less.

The start-up time of the entire plant can be shortened by coordinating the boiler and the steam turbine, but the above-mentioned conventional methods are all rapid start-up methods focusing on only one of the boiler and the steam turbine. Even if such individual methods are combined, there is no guarantee that the startup time of the entire plant will be the shortest. This is because the boiler and the steam turbine have very strong mutual interference, and the individual optimization does not necessarily become the overall optimization. The same applies to a combined cycle power plant consisting of a gas turbine, an exhaust heat recovery boiler, and a steam turbine, and the individual operations depend on the NOx emissions, the output of the exhaust heat recovery boiler and the steam turbine, and the generated thermal stress due to mutual interference of equipment. Affect.

Therefore, by using a dynamic characteristic simulator simulating the basic characteristics of the entire system, humans input driving patterns such as various start-ups in advance while adjusting them, evaluate them, and then apply them to an actual machine. The method of doing is widely used.

In addition, the operation schedule is not adjusted by a human, but the objective evaluation function (for example, in the case of start-up, the time required for start-up) is minimized while considering various operation limiting conditions in the dynamic characteristic simulator. It is also conceivable to present the operation plan by the optimizing technique to be realized.

There are many optimization techniques such as the maximum gradient method, the dynamic programming method, the fuzzy control method, the neuro-construction method, etc., but various devises have been made to apply them to individual problems. In either method, the dynamic characteristic simulator is repeatedly used.

[0011]

However, in any of the above methods, it is difficult to actually obtain an optimum solution because of the capacity limit of the optimization technique and the practical limit due to the speed of the computer used. , Greatly simplify the problem (for example, reduce the number of manipulated variables), or simplify the dynamics simulator (ignore complex phenomena or approximate static characteristics) and get an answer in a short time Either or both are used.

As described above, in the conventional method, whether the object is limited to the plant part or the whole system is handled, the problem is simplified or the simulator to be used and the operation restriction condition are simplified. There was a side that could not get the conclusion and judgment directly connected to the plant.

Although the above is a conventional example of optimization of the starting process, the same technique can be applied to the stopping process, the load change, and the treatment process in case of abnormality or accident, and the same technique can be applied. I have a problem.

The problem to be solved by the present invention is that the heat generated in the boiler and the steam turbine is difficult to realize in the operation control system of the power generation plant such as thermal power generation due to the limitation of the conventional method and the capacity limitation of the computer used. The objective is to enable the automatic creation of an optimal operation schedule that minimizes the time required for starting and stopping while simultaneously satisfying the operating restriction conditions regarding stress and NOx emissions and the environmental regulation value in a form close to the actual conditions.

The present invention is also intended to further shorten the time required to create such an optimum operation schedule within the range of the capacity of the computer used and within the range of practical labor. It is a task to do.

Further, at the time of start-up in an actual plant, it is possible to make an optimum correction when there is an instruction to change operating conditions such as environmental regulation values and start-up completion time from a central power supply command station (a place to monitor the power supply situation in the whole area). Doing so and increasing the efficiency of the startup test are also problems to be solved by the present invention.

[0017]

In order to solve the above problems, the present invention is provided with the following four kinds of means (A) to (D) as shown in FIG.

(A) Prediction accuracy of a plant state with respect to various operation settings requires a calculation time, but has a very high dynamic characteristic simulator and a function capable of reflecting and evaluating the operation restriction conditions of an actual machine in detail. Predictor.

(B) As compared with the first predicting means, the speed of predicting the plant state with respect to the operation setting is inferior in accuracy, but a faster dynamic characteristic simulator and a function capable of evaluating operation restriction conditions at higher speed and in a simple manner. A second predicting means including and.

(C) When the second predicting means is used,
Optimizing means (optimum searching means) for automatically searching the optimum operation setting contents of plant operation for at least one event at the time of stop, load change, and accident.

(D) A result obtained by inputting the optimum operation setting content obtained by the optimizing means to the first predicting means and a result of the second predicting means with the optimum operation setting content. The difference is evaluated, and the second predicting means is instructed to correct the operation limiting condition, the model of the dynamic characteristic simulator and the input constant, and further change the operation setting content, and again the optimizing means. Optimal operation application adjusting means having a function of repeating a series of procedures for performing optimization until the result of the first predicting means satisfies the operation limiting condition of the actual equipment and can be applied to the actual equipment.

These means are realized as software functions of a computer. Along with this, the input means of the actual machine data, the display means for the operator, etc., and the output means for the control equipment of the actual machine plant are both attached to the hardware and the software.

Further, the contents of the second prediction means of (B) are
It consists of the following sub-means. That is, the operating schedule assumption means for preliminarily estimating the operating schedule and the operating characteristics when the plant is operated according to this operating schedule are predicted, that is, thermal stress generated in the boiler and the steam turbine and NOx from the boiler. A high-speed predicting means for quantitatively calculating the emission amount, an operating characteristic evaluating means for evaluating the thermal stress characteristics and the NOx characteristics obtained by this means by comparing them with operating limiting conditions, and based on the operating characteristic evaluation results And an operation schedule correction means for correcting the above-mentioned assumed operation schedule in order to obtain better operation characteristics, and a process of repeating the above-mentioned operation schedule assumption, plant dynamic characteristic prediction, operation characteristic evaluation, and operation schedule correction. Optimality determination means for determining whether or not the operation schedule has converged to the optimum value Has a operation schedule setting means for setting the obtained optimum operation schedule to the device control system, the operation schedule display means for presenting to the operator via the display device the operation schedule.

In the present invention, each of the above means operates as follows. It is assumed that the dynamic characteristic simulators installed in the first predicting means and the second predicting means have a speed difference of tens to hundreds of times or more.

First, an operation schedule that is initially assumed such as startup is given to the second predicting means. The initially assumed operation schedule is an operation schedule that is assumed to have the stop time before activation and the temperature state of the device as initial conditions.
Instrument temperature is automatically collected. The simplified plant dynamic characteristic model can quantitatively predict the characteristic when the plant is operated according to the assumed start-up schedule based on the physical model. The "initially assumed operating schedule" is the initially assumed schedule, while the "assumed start-up schedule" is the previously obtained schedule in each iteration process in the inner loop. Say. Both will be the same on the first iteration. By this quantitative prediction, not only the response of the main temperature, pressure, flow rate, etc. of the plant is known, but also the thermal stress and NOx emission amount of the boiler and steam turbine as the plant state are the operation limiting conditions by the characteristic evaluation means. You can check whether or not It is desirable that the schedule correction means that functions when not satisfied should improve the operation schedule by a method similar to the way of thinking of operators and designers. Therefore, for example, fuzzy reasoning is applied as the optimizing means, and the driving characteristic obtained above is converted into a weighted qualitative evaluation result to correct the driving schedule. The fuzzy rule used at this time is about the relationship between each operation characteristic at startup and the schedule correction amount corresponding to each operation,
It is created based on the qualitative causal relationship information that the operator and the designer have as knowledge. The optimization means also determines whether or not the operation schedule has converged to the optimum value in the process of repeating the above-mentioned operation schedule assumption, plant dynamic characteristic prediction, operation characteristic evaluation, and operation schedule correction (this is called an optimum value search step). As a result of the convergence, under the conditions given to the second predicting means, the time required for the operation such as the start-up time is short, the safety in plant operation is high, and the optimal operation schedule which is environmentally friendly can be obtained.

The operation schedule result thus obtained is input to a simulator in the first predicting means capable of evaluating the state of the plant in further detail to make a prediction, and in a form closer to an actual machine, thermal stress and environmental compatibility The operation restriction condition including the property is evaluated by the result of the first prediction means.

In general, when the optimum operation schedule obtained by the second predicting means is evaluated by the first predicting means due to the difference in the accuracy of the simulator and the difference in the evaluation accuracy of the operation restriction conditions. Is not always optimal and exceeds the driving restriction condition, or has too much driving margin, and the state quantity not evaluated by the simulator in the simplified second prediction means is limited. It may occur that the value exceeds.

Therefore, the optimum operation adjusting means compares the conditions and the results of the first predicting means and the second predicting means with each other to reduce the difference between them.
The adjustment means of the operation constraint condition and the judgment reference value in the prediction means, and the input condition of the simulator and the input value of the model are adjusted. This adjustment is also effective by iterative learning, and fuzzy algorithm, neuro algorithm, steepest gradient method, etc. can be applied as means.

As described above, the present operation plan preparation apparatus has the inner iterative convergence loop that passes through the second predicting means and the optimizing means, and the outer iterative convergence loop that passes through the first predicting means and the optimum operation application adjusting means. It consists of a loop and a double loop.

Further, the finally provided operation schedule at the time of start-up is set in the equipment control system to control the operation of the actual plant. As the start-up schedule, the CRT display data for presenting the optimum value search process and the optimum start-up schedule obtained as a result thereof to the operator are edited and the start-up schedule is transferred to the CRT display device.

By the action of each means in the present invention described above, the operation limiting conditions such as the thermal stress and NOx emission amount appearing as a result of the interference between the equipments constituting the plant are satisfied, and the starting and stopping processes are performed. An operation plan that minimizes the time required is shown. With this configuration, the optimum operation schedule of the plant applicable to the actual machine can be provided by overcoming the limit of the realistic optimization algorithm and the limit of the simulation capability by the computer.

It is also possible to create an optimum start-up schedule at the power station site that can complete start-up, stop, etc. on time specified by the central power supply command station. Furthermore, even when the operation limit value is changed during plant operation, each means described above operates in the same way, and an optimum operation schedule that matches the new operation limit value after change is automatically generated and executed. It becomes possible to do.

On the other hand, when the expected performance cannot be obtained in the start-up test, the characteristics may be adjusted while conducting repeated tests several times for sensitivity analysis. In such a situation, in order to determine the operation schedule in the next test, the actual plant itself was used instead of the first dynamic simulator of the outer loop, and the test was performed more accurately for each test. It is also possible to evaluate the driving method in advance.

[0034]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, as an embodiment of the present invention, an operation control system for a combined cycle power plant including a gas turbine, an exhaust heat recovery boiler and a steam turbine will be described. Take up and explain.

FIG. 2 shows a basic configuration of a combined cycle power plant 300 and an operation control system 1000 to which the present invention is applied.

The operation control system 1000 is roughly divided into an optimum operation schedule search means 100, an equipment control system 200, and an optimum operation application adjusting means 400. The optimum driving schedule search means 100 is further composed of a start schedule assumption means 110, a high speed prediction means 500, a driving characteristic evaluation function 120, an optimality evaluation function 140, a start schedule correction means 150, a fuzzy rule 160, and a start schedule setting means 170. Has been done. The high-speed predicting means 500 is shown outside the frame of the optimum driving schedule searching means 100 in FIG. 2 for convenience of illustration. Other,
The operation control system 1000 also has a startup schedule display unit 180 and a CRT display device 20.

The optimum operation application adjusting means 400 comprises a high precision predicting means 410, an actual machine application determining section 430 and a total optimization adjusting section 420. The equipment control system 200 for controlling the equipment of the entire system includes a gas turbine control system 210, a steam turbine control system 220, and an HRSG control system 230. In addition, combined cycle power plant 3
00 is roughly divided into a gas turbine facility 320 and a heat recovery steam generator (HRSG: Heat Recovery Steam Generato).
r) 310 and steam turbine equipment 330.

When this configuration is made to correspond to FIG. 1, the high-precision predicting means 410 corresponds to the first predicting means, and the high-speed predicting means 500 and the driving characteristic evaluation function 120 together correspond to the second predicting means. . Therefore, the prediction unit 410 is a general term having a function of evaluating all the limiting conditions such as thermal stress. In addition, the inner optimum adjustment units are 140, 150, 160.
And the outer optimum adjusting units correspond to 400, 430, and 420. Therefore, the inner loop is 111, 531,
The route corresponds to the route passing through 121 and 151, and the outer loop corresponds to the route from 171 to 421 through the functional blocks of 410, 430 and 420.

Here, the operation control system 100 of the present invention.
Before explaining 0, the combined cycle power plant 3
A description will be given of the operation principle centering on the startup of 00.

In the gas turbine equipment 320, the fuel 322
Is injected into the combustor 323, and the combustion air 324 is press-fitted by the compressor 329, whereby the energy generated by the combustion is converted into mechanical energy by the gas turbine 326, whereby the generator connected to the common shaft 328. 335 is driven and converted into electric energy, and a part thereof becomes driving force of the compressor 329. At the time of startup, the fuel control valve 321 is operated according to the fuel control valve opening command 211 from the gas turbine control system 210 to control the fuel flow rate, so that the gas turbine equipment 320 and the common shaft 328 are controlled.
The steam turbine equipment 330 and the generator 335 connected to the are accelerated.

Also, the exhaust gas 3 from the gas turbine 326
27 is led to the exhaust heat recovery boiler equipment 310, and exhaust gas 32
The thermal energy of 7 is recovered. At this time, the exhaust gas 327 causes heat to evaporate and superheat the fluid in the various heat exchangers arranged in the smoke tunnel 314 of the exhaust heat recovery boiler facility 310. The exhaust heat recovery boiler equipment 310 of the present embodiment comprises a steam system having three pressure levels, and the steam generated from each is a high pressure steam 241, a medium pressure steam 242, and a low pressure steam 243. Due to the thermal energy of these steams, in the steam turbine equipment 330, the high pressure turbine 331 and the intermediate pressure turbine 33, respectively.
2. The low-pressure turbine 333 is driven, and plays a role in power generation by the generator 335 connected to the common shaft 328. At startup, the above-described high-pressure main steam 241, intermediate-pressure main steam 242, and low-pressure main steam 243 generated from the exhaust heat recovery boiler facility 310.
The high pressure bypass valve 234 and the medium pressure bypass valve 23, respectively.
5. Each pressure is controlled to a predetermined value by bypassing through the low pressure bypass valve 236, and the high pressure control valve 231, the intermediate pressure control valve 232, and the low pressure control valve 233 are opened to operate the high pressure turbine 331, The outputs of the pressure turbine 332 and the low pressure turbine 333 are increased. Therefore, the high pressure bypass steam 244, the medium pressure bypass steam 24
5. The flow rate of the low-pressure bypass steam 246 increases as the openings of the bypass valves 234, 235, 236 increase, and the flow rates of steam into the steam turbine increase as the openings of the regulator valves 231, 232, 233 increase. And decrease by that amount. The opening commands 221, 222 to these bypass valves,
223 and opening / closing commands 224, 225, 226 to the regulator valve
Are both output from the steam turbine control system 220. Further, when the temperature deviation between the medium-pressure superheated steam 301 and the high-pressure turbine exhaust 302 falls within a predetermined value during the startup of the plant, the medium-pressure stop valve 237 is opened. At this time, the operation signal 227 is also commanded by the steam turbine control system 220. On the other hand, the HRSG control system 230 controls the water level inside the steam drums 306, 307, 308,
It has a function of spraying water into the pipes 242 and 243 so that the temperature of the steam in the pipes 242 and 243 does not rise excessively (however, in FIG. 2, the signal and the operating end are not shown). Condensed water 337 from the condenser 334 is supplied to the low-pressure water supply pump 316, the medium-pressure water supply pump 317, and the high-pressure water supply pump 318.
The low pressure drum 306 and the medium pressure drum 30 respectively.
7. To keep the water level of the high-pressure drum 308 within a predetermined value,
Low-pressure water supply 303, medium-pressure water supply 304, high-pressure water supply 3
The flow rate is controlled as 05. As described above, this control is performed by adjusting the water level set point together with the HRSG control system 2
30 is in charge.

Here, the factors that limit the operation when the plant state changes slowly over a relatively long time such as start-up are the high-pressure turbine 331 and the intermediate-pressure turbine 3.
32, the thermal stress generated in the rotor, the thermal stress generated in the header of the header of the high-pressure superheater 311 and the medium-pressure superheater 312 at the thin-tube outlet, and the smoke tunnel outlet 315 of the exhaust heat recovery boiler facility 310 to the atmosphere. It is the amount of NOx emissions. Each of the thermal stresses described above is a dynamic process with a large time delay, such as heat transfer from the gas turbine exhaust gas 327 to the metal of the heat exchanger, heat transfer from the metal to the internal fluid, and heat transfer from the internal fluid to the metal of interest. Appears as a result of. The NOx emission amount from the boiler also largely depends on the NOx emission characteristic of the gas turbine itself and the temperature characteristic of the denitration device 313 installed in the smoke tunnel.
Therefore, in order to accurately manage these operation limiting factors, the coordination and consistency of the above-mentioned control operations are required.

The operation control system 10 of the combined cycle power plant 300 will be described below with reference to FIG.
The operating principle of 00 is outlined.

As a start request 11 to the plant, the central power supply command station or the operation planner 10 commands the operation control system 1000 about the target start completion time and the target load.
In response to this, the startup schedule assumption unit 110 of the optimum operation schedule search unit 100 creates a temporary startup schedule 111 according to the time when the plant was stopped and the equipment temperature, and sends it to the high speed prediction unit 500. . In the high-speed predicting means 500, the provisional activation schedule 111 is determined by the dynamic characteristic model 532 of the device control system.
Along with gas turbine model 533, boiler model 53
4. Start control of the steam turbine model 535. The device control at this time is performed by a method similar to that described in the actual machine as much as possible, and the dynamic characteristic 531 of the plant is obtained as if the plant was started up, and is sent to the next operating characteristic evaluation function 120. For details of the model, see "Dynamic Simulation of an Ad" by T. Akiyama et al.
vanced Combined Cycle Plant with Three Pressure an
d Reheat Cycle ", Proceedings of ICOPE'93 (JSME-ASM
E International Conference on Power Engineering- 9
3), pp.221-226. The operating characteristic evaluation function 120 further includes a turbine thermal stress characteristic evaluation function 122 for evaluating the thermal stress generated in the rotor of the high-pressure and intermediate-pressure steam turbine, and a high-pressure and intermediate-pressure superheater outlet of the boiler. Boiler thermal stress characteristic evaluation function 123 for evaluating thermal stress generated in header, exhaust heat recovery boiler 31
It comprises an exhausted NOx characteristic evaluation function 124 for evaluating NOx exhausted from 0, and a startup required time evaluation function 125. The operating characteristic evaluation function 120 includes these functions 12
The values obtained from Nos. 2, 123, and 124 are compared with the operation restriction conditions for each, and the margin values for them are evaluated. This is for calculating the time required. When these evaluations are completed, the processing moves to the next optimality evaluation function 140. Here, it is for determining the optimality of the assumed startup schedule, and the optimal startup schedule having the shortest startup time is identified while the various evaluation results of the startup characteristics satisfy the operation restriction conditions. Therefore, normally, the optimum startup schedule is not determined at the first time. In order to improve the activation schedule based on the activation characteristic evaluation result 121, the next activation schedule correction means 150 together with the activation characteristic evaluation result 121.
Processing is passed to. The start-up schedule correction means 150 corrects the start-up schedule by applying fuzzy inference based on the evaluation result obtained by the driving characteristic evaluation function 120. Start schedule modification amount 1 determined here
51 is transferred to the activation schedule assumption means 110, and the activation schedule is assumed again.

The fuzzy rule 160 used here is composed of a turbine thermal stress adjustment rule 162, a boiler thermal stress adjustment rule 163, and an exhaust NOx adjustment rule 164, and an expert has knowledge as to the relationship between the startup characteristic and the startup schedule correction amount. It is created based on qualitative causal relationship information.

In the process of repeating the above-mentioned start-up schedule assumption, plant dynamic characteristic prediction, start-up characteristic evaluation, and start-up schedule correction (this is called an optimum value search process), the optimumness evaluation function 140 determines the optimum start-up schedule. It is determined whether or not the value has converged. As a result of the convergence, an optimal startup schedule that has a short startup time and satisfies the operation restriction condition can be obtained. When the optimal activation schedule is determined by the optimality evaluation function 140, a schedule parameter (that is, the optimal operation schedule of the inner loop) 171 that defines this optimal activation schedule is sent to the optimal operation application adjusting means 400 via the activation schedule setting means 170. Is set.

In the optimum operation application adjusting means 400, the controller part 412 which inputs the optimum operation schedule 171 of the inner loop and determines the control operation of the high precision simulator.
A more detailed reproduction simulation is performed by using the physical model 411 that simulates the characteristics of the main body such as GT, HRSG, and ST. The result is the actual device application determination unit 430.
Then, it is determined whether or not it is okay to directly apply the schedule instructed in 171 to the actual machine by comparing the actual detailed operation restriction conditions with. That is, if the state of the plant does not exceed the allowable value, the schedule is passed to the device control system 220 to be applied to the actual machine as it is. If the state exceeds the allowable value, the total optimization adjusting unit 420 recognizes and stores the difference between the results of the simulator of the high speed prediction means 500 and the simulator of the high accuracy prediction means 410 when the schedule parameter 171 is obtained. However, the parameter setting values of the high-speed prediction means 500, the setting limit values of the operating characteristic evaluation function 120, and the optimality evaluation function 14
The judgment criterion value of 0 and the optimization rule of the fuzzy rule 160 are reviewed, and in the next simulation, the result of the high precision prediction means 410 is in the direction of falling within the limit value. Modify and set the parameters. (This correction will be mainly given to the high-speed prediction means 500.) The search process of the optimum start schedule is displayed on the CRT display device 20 via the start schedule display means 180.
At this stage, the operator can be presented with the search situation and the optimum start schedule which is the convergence result. The display contents in this case are the assumed start-up schedule, the dynamic characteristic prediction result in each of the high-speed predicting means 500 and the optimum operation application adjusting means 400, the margin value for the operation limit value, the start-up required time, the difference between both, and the like. .

The outline of the embodiment in which the plant operation plan creation device according to the present invention is applied to a combined cycle power plant has been described above. Hereinafter, the present embodiment will be described more specifically.

FIG. 3 shows the optimum operation schedule search means 1
00 shows the basic idea of the activation schedule optimization. A gas turbine start-up plan and a steam turbine start-up plan are executed based on the plant dynamic characteristic prediction values obtained from the high-speed predicting means 500, based on the thermal stresses of the boiler and the steam turbine and the exhaust NOx characteristics. At this time, in the gas turbine startup plan, the gas turbine main plan (GTPS: Gass Turbine Primal Scheduling
Abbreviated) and thermal stress in consideration of gas turbine wide area adjustment (GT
GT: Gas Turbine Global Tuning) was performed, and in the steam turbine start-up plan, steam turbine main plan (STPS: Abbreviation of Steam Turbine Primal Scheduling) considering thermal stress and steam turbine local adjustment (S
TLT: Steam Turbine Local Tuning). Since the starting method of the gas turbine has a great influence directly on the exhausted NOx characteristic, the GTPS has a function of creating a finely detailed startup schedule of the gas turbine according to the predicted value of the exhausted NOx characteristic. Further, since the starting method of the steam turbine has a great influence directly on the thermal stress characteristics, in STPS, the starting schedule of the steam turbine is made to be finely detailed as a whole according to the predicted value of the thermal stress characteristics. On the other hand, thermal stress is indirectly influenced by the starting method of the gas turbine via the heat transfer of the exhaust heat recovery boiler. Therefore, in GTGT, the starting schedule of the gas turbine is finely divided according to the predicted value of the thermal stress characteristics. Have a function of adjusting. In addition, since the characteristics of the denitrification equipment are indirectly changed as a result of the change in the heat absorption characteristics of the exhaust heat recovery boiler due to the steam turbine start-up method, the emission NO
Partially fine-tunes the steam turbine startup schedule according to the predicted value of x-characteristic.

Next, referring to FIG. 4, an example of a parameter (hereinafter, referred to as a schedule parameter) that defines the plant start schedule will be described in relation to the plant start process.

As shown in FIG. 4, as the schedule parameters related to the gas turbine, the speed increase rate (DN), rated speed holding time (DTNL), initial load (LI), initial load holding time (DTLI), first load Rate of increase (DL1), load holding time (DTH
L), 2nd load increase rate (DL2), 3rd load increase rate (DL3)
It is. The gas turbine is started by adjusting the opening degree of the fuel control valve 321 that is the operating end with the start schedule defined by these as the control target. Also,
As already mentioned as the operating end of the steam turbine,
High pressure bypass valve (HPBV), medium pressure bypass valve (IPBV), low pressure bypass valve (LPBV), high pressure control valve (HPCV), medium pressure control valve (IPCV), low pressure control valve (LPCV), medium pressure stop valve (ISHV) ), And control according to the following schedule parameters.
That is, as the schedule parameters, the high pressure bypass valve operating speed (DAHBV), the medium pressure bypass valve operating speed (D
AIBV), low-pressure bypass valve operation speed (DALBV), low-pressure bypass valve operation waiting time (DTLBV), high-pressure control valve first operation speed (DAHCV1), high-pressure control valve second operation speed (DAHCV2), low-pressure control valve operation speed (DALCV). The control targets and operation timings other than these schedule parameters are as illustrated. Where TMS as "steam condition"
Is the high-pressure main steam temperature, PMS is the high-pressure main steam pressure, PCRP is the high-pressure turbine exhaust pressure, PIS is the medium-pressure main steam pressure, PLS is the low-pressure main steam pressure, ΔT is the deviation between the high-pressure superheated steam temperature and the high-pressure turbine exhaust temperature. is there. ata represents kg / cm 2.
Further, in the figure, the% display of each valve indicates the opening.

Next, the optimization search means 100 is used.
A high speed prediction means 500 for predicting thermal stress and exhaust NOx of a steam turbine and a boiler will be described.
First, the thermal stress of a steam turbine is calculated according to the above-mentioned US Pat.
It was modeled by the method described in detail in No. 8,359. That is, the steam condition (temperature, pressure) inside the steam turbine and the heat transfer coefficient on the rotor surface are estimated from the steam condition at the inlet of the steam turbine, the speed, and the load, and the unsteady temperature distribution inside the rotor metal is calculated, This is a method of calculating the thermal stress of the bore. This is the high pressure turbine 331 and the intermediate pressure turbine 3 that should be managed as a limiting condition when the plant is started.
Applied to 32 thermal stress calculations. The thermal stress of the boiler was modeled by the method described in detail in JP-A-59-157402. That is, the internal heat transfer coefficients of the high-pressure superheater outlet header and the medium-pressure superheater outlet header, which should be noted as limiting conditions, are estimated from the steam conditions (temperature, pressure) and the flow rate, and the internal metal transfer rate of The unsteady temperature distribution was calculated to calculate the thermal stress on the inner and outer surfaces of the header metal.

Emission NOx is predicted by using a dynamic characteristic model. This model is composed of a denitration device model and a denitration control system model. The denitration device model is an NOx exhausted to the atmosphere by the inlet gas temperature TG, the gas flow rate GG, the inlet gas NOx concentration P1 and the ammonia injection amount GNH3 from the denitration control system model. Calculate the instantaneous value PS. This value is converted by the moving average means, for example, as a moving average value PA per hour.

Next, the turbine thermal stress characteristics, the boiler thermal stress characteristics, and the exhaust NO obtained from the high speed predicting means 500.
The operating characteristic evaluation function 120 for evaluating the x characteristic (in the case of this description, the thermal stress at the time of startup, the relationship with operating limiting conditions such as NOx, etc. is evaluated) will be described.

FIG. 5 shows a turbine thermal stress characteristic evaluation function 12
2 shows the evaluation method of the turbine thermal stress characteristics in 2. First, the time zone (t1 to t6) from the start of the turbine to the elapse of a predetermined time after the start of the turbine is divided into a plurality of sections (in this example, 5 divisions). Then, the minimum thermal stress margin m (i) in the i-th section is obtained. Where the thermal stress margin m
Is the limit value SL and the calculated value by the dynamic characteristic model is S (kg / mm
2), it is defined by the following formula.

[0056]

[Equation 1] m = S _L −S ------------------------- (1) As already mentioned, in fact, attention is paid to thermal stress. Since there are four places, the surface and bore of the high-pressure turbine and the surface and bore of the medium-pressure turbine, there are four minimum thermal stress margins to be obtained for each section, and these are given below.

MHS (i): minimum thermal stress margin of high pressure turbine surface in section i mHB (i): minimum thermal stress margin of high pressure turbine bore in section i mIS (i): minimum thermal stress margin of intermediate pressure turbine surface in section i mIB (i): The minimum thermal stress margin of the medium pressure turbine bore in the section i is also evaluated by the thermal stress characteristic evaluation function 123 of the boiler.
Basically, it is similar to the case of the turbine thermal stress characteristic evaluation function 122. However, the thermal stress points to be noted are two points on the inner surface of the header of the high pressure superheater and the intermediate pressure superheater. This is because in the case of the header, the thermal stress generated on the outer surface is smaller than that on the inner surface, and therefore it is sufficient to focus only on the inner surface. Therefore, there are two minimum thermal stress margins to be obtained for each section, which are as follows.

MHHD (i): Minimum thermal stress margin on the inner surface of the high pressure superheater header in the section i mIHD (i): Minimum thermal stress margin on the inner surface of the medium pressure superheater header in the section i Next, a method for evaluating exhausted NOx characteristics will be described. .

FIG. 6 shows an exhaust NOx characteristic evaluation method in the exhaust NOx characteristic evaluation function 124. Similar to the turbine thermal stress characteristic evaluation function 122, this method first divides the time zone (t1 to t7) from the start of gas turbine startup to the elapse of a predetermined time after completion of plant startup into multiple sections.
(In this example, the case of 6 divisions is shown), and the minimum emission NOx instantaneous value margin mPS (i) and the minimum emission NOx average value margin mPA (i) in the i-th section are obtained. Here, the exhausted NOx instantaneous value margin mPS and the exhausted NOx average value margin mPA are the respective limit values calculated by PSL, PAL, and the dynamic characteristic model.
PS and PA are defined by the following equation.

[0060]

[Equation 2] m _PS = P _SL −P _S ---------------- (2)

[0061]

[Equation 3] m _PA = P _AL −P _A ---------------- (3) FIG. 7 is for correcting the startup schedule by using the evaluation result of the above dynamic characteristics. 3 shows the relationship between the fuzzy rule 160 and the modification schedule parameter. In the embodiment shown in the figure, four parameters (DAHBV, DAIBV, DAHCV1, DAHCV2) considered to be particularly effective in the steam turbine relation are targeted as the correction schedule parameter, and eight parameters (in the gas turbine relation are D
N, DTNL, LI, DTLI, DL1, DTHL, DL2, DL3). The turbine thermal stress adjustment rule 162 and the boiler stress adjustment rule 163 are used in both the steam turbine main plan (STPS) and the gas turbine wide area adjustment (GTGT).
The emission NOx adjustment rule 164 is the steam turbine local adjustment (S
Used for TLT) and Gas Turbine Master Plan (GTPS). In addition, the parameters to be corrected by STPS, GTGT, STLT, and GTPS are indicated by a circle in the figure.

FIG. 8 shows a membership function used in the fuzzy rule 160. This function is determined based on the plant dynamic characteristics obtained by the operating characteristic evaluation function 120. FIG. 8 (1) shows the membership function for STPS, and (a) is the membership function for evaluating the thermal stress margin used in the conditional part of the fuzzy rule, and has four functions (NS, ZO, PS, PB). (B) is a membership function for modifying the steam turbine related schedule parameters used in the conclusion part, and includes five functions (NB, N
S, ZO, PS, PB). Here, the meaning of each membership function is NB: Negative Big, NS:
Negative Small, ZO: Zero, PS: Positive Small,
PB: Positive Big. FIG. 8 (2) shows the membership function for GTGT, and (a) is the membership function for evaluating the thermal stress margin used in the conditional part of the fuzzy rule, and has four functions (NS, ZO, PS, PB). And (b) is a membership function for modifying the gas turbine related schedule parameter used in the conclusion part, which is composed of three functions (NS, ZO, PS). The membership functions for thermal stress margin evaluation in (1) and (2) above are commonly applied to both turbine thermal stress and boiler thermal stress. FIG. 8 (3) shows the membership function for GTPS, and (a) and (b) are the membership function for the emission NOx instantaneous value margin evaluation and the emission NOx average value margin used in the conditional part of the fuzzy rule, respectively. There are four membership functions for evaluation (NB, NS, Z).
O, PS) and (NS, ZO, PS, PB). Further, (c) is a membership function for modifying the gas turbine-related schedule parameter used in the conclusion part,
It consists of 5 functions (NB, NS, ZO, PS, PB).
FIG. 8 (4) shows the membership function for STLT, and (a) and (b) are the membership function for the emission NOx instantaneous value margin evaluation and the emission NOx average value margin used in the conditional part of the fuzzy rule, respectively. It is a membership function for evaluation, and each has four functions (NB, NS, ZO,
PS) and (NS, ZO, PS, PB). Further, (c) is a membership function for modifying the steam turbine-related schedule parameter used in the conclusion part, which is composed of five functions (NB, NS, ZO, PS, PB).

Next, a fuzzy rule using the above membership function will be described.

FIG. 9 shows the entire structure of the fuzzy rule 160. As described above, the fuzzy rule 160 is roughly divided into a turbine thermal stress adjustment rule 162, a boiler thermal stress adjustment rule 163, and an exhaust NOx adjustment rule 164. Become.

Turbine thermal stress adjustment rule 162 is ST
It is divided into PS and GTGT. For STPS, it is for modifying the start schedule parameter of the steam turbine for the purpose of adjusting turbine thermal stress.
For GTGT, it is for modifying the start schedule parameter of the gas turbine for the purpose of adjusting the turbine thermal stress. As shown in FIG. 9, both are further composed of four types of rule tables (HS, HB, IS, IB). The HS rule table is intended to adjust the thermal stress on the surface of the high pressure turbine rotor.
Four rule tables (HS12, HS23,
HS34, HS45). The HB rule table is for adjusting the high pressure turbine rotor bore thermal stress, the IS rule table is for adjusting the medium pressure turbine rotor surface thermal stress, and the IB rule table is for adjusting the medium pressure turbine rotor bore thermal stress. Is intended for
Each consists of four types of rule tables.

The boiler thermal stress adjustment rule 163 is STP.
It is divided into S and GTGT. STPS is for modifying steam turbine start-up schedule parameters for the purpose of adjusting boiler thermal stress, and GTGT is for starting gas turbine for the purpose of adjusting boiler thermal stress. It is for modifying the schedule parameters. Further, it is composed of two types of rule tables (HHD, IHD). The HHD rule table is for adjusting the high-pressure header thermal stress, the IHD rule table is for adjusting the medium-pressure header thermal stress, and is further composed of four types of rule tables.

The emission NOx adjustment rule 164 is GTPS.
NOx emission for GTPS and for STLT
For correcting the start schedule parameter of the gas turbine, and for STLT for correcting the start schedule parameter of the steam turbine for the purpose of adjusting exhaust NOx. Both of them are further composed of a rule table (PS, PA).
The PS rule table is for the purpose of adjusting the instantaneous value of exhausted NOx, and the PA rule table is for the purpose of adjusting the average value of exhausted NOx, and is further divided into five types of rule tables.

Next, the contents of the rule table will be specifically described with a representative example.

FIG. 10 shows STPS, GTGT, and GT.
A concrete example of the rule table for PS and STLT is shown, and the STPS-HS rule table of FIG. 9 is shown as a typical example. Here, steam turbine main plan (ST
As part of the turbine thermal stress adjustment rule used in PS), the STPS-HS rule table for adjusting the minimum thermal stress margin of the high-pressure turbine rotor surface is illustrated. In this table, the two numbers that follow HS, such as HS23, indicate the numbers of the two types of time intervals (in this example, the second and third) in which the impact is considered. Here, high-pressure bypass valve operating speed (DAHBV), medium-pressure bypass valve operating speed (DAIBV), high-pressure adjusting valve first operating speed (DAHCV1), high-pressure adjusting valve second operating speed as four correction schedule parameters.
A fuzzy rule for correcting (DAHCV2) in relation to the minimum thermal stress margin of the high-pressure turbine rotor surface in the five sections of interest shown in FIG. 5 is shown. That is, STP
In the S-HS12 rule table, the high pressure turbine rotor surface minimum thermal stress margin in the first section and the second section
DAHBV and DAH from the combination of qualitative relations of mHS (1) and mHS (2)
It defines the amount of modification of CV1. In the STPS-HS23 rule table, from the combination of the qualitative relationships of the margins mHS (2) and mHS (3) in the second section and the third section, the same
It defines the correction amount for DAHBV and DAHCV1. STPS-H
In the S34 rule table, the correction amounts of DAHBV, DAHCV1, and DAHCV2 are defined by the combination of the qualitative relationships of the margins mHS (3) and mHS (4) in the third section and the fourth section.
In the STPS-HS45 rule table, the correction amount of DAHCV2 is defined by the combination of the qualitative relations of the margins mHS (4) and mHS (5) in the fourth section and the fifth section. However, the blank part of the rule table means that the causal relationship between each thermal stress margin and the schedule parameter is small or almost nonexistent.

As a part of the turbine thermal stress adjustment rule used in the gas turbine global adjustment (GTGT), the HS rule table for adjusting the minimum thermal stress margin of the high pressure turbine rotor surface is 6 based on the same concept as in FIG. Acceleration rate (DN), initial load holding time (DTLI), first load rising rate (DL1), load holding time (DTHL), second load rising rate (DL) as one correction schedule parameter
2) Define a fuzzy rule for correcting the third load increase rate (DL3) in relation to the minimum thermal stress margin of the high pressure turbine rotor surface in the five target sections shown in FIG. That is, in the HS12 rule table, the correction amounts of DN, DTLI, and DL1 are defined by the combination of the qualitative relations of the minimum thermal stress margins mHS (1) and mHS (2) of the high-pressure turbine rotor surface in the first section and the second section. ing. In the HS23 rule table, the DTL is calculated by combining the qualitative relations of the margins mHS (2) and mHS (3) in the second section and the third section.
It defines the amount of modification of I, DL1, and DTHL. In the HS34 rule table, the margin in the third section and the fourth section
From the combination of qualitative relationships of mHS (3) and mHS (4), DL1, DTH
It defines the correction amount of L and DL2. In the HS45 rule table, the margin mHS in the 4th section and the 5th section
(4), mHS (5) from the combination of qualitative relationships, DTHL, DL
2. It defines the amount of modification of DL3. Also in this case, the blank part of the rule table means that the causal relationship is small or almost nonexistent, as in the previous case.

As a part of the emission NOx adjustment rule used in the gas turbine main plan (GTPS), in the PS rule table for adjusting the minimum emission NOx instantaneous value margin, the speed increase rate (DN) as the eight correction schedule parameters, Rated speed holding time (DTNL), initial load (LI),
The initial load holding time (DTLI), the first load increasing rate (DL1), the load holding time (DTHL), the second load increasing rate (DL2), and the third load increasing rate (DL3) are shown in FIG. The fuzzy rule for correcting in relation with the minimum discharge NOx instantaneous value margin in the section of interest is shown. That is, in the PS12 rule table, DN, DTNL, LI, DTLI, DL1, DTHL, DL2 are obtained from the combination of the qualitative relations of the minimum emission NOx instantaneous value margins mPS (1) and mPS (2) in the first section and the second section. ,
It defines the modification amount of DL3. In the PS23 rule table, the margin mPS (2) in the second section and the third section,
From the combination of qualitative relationships of mPS (3), DTNL, LI, DTLI,
It defines the modification amount of DL1, DTHL, DL2, DL3. PS3
In the 4 rule table, from the combination of the qualitative relations of the margins mPS (3) and mPS (4) in the third section and the fourth section,
It defines the amount of modification of LI, DTLI, DL1, DTHL, DL2, DL3. In the PS45 rule table, the correction amounts of DTHL, DL2, and DL3 are defined by the combination of the qualitative relations of the margins mPS (4) and mPS (5) in the fourth section and the fifth section. P
In the S56 rule table, the correction amounts of DTHL, DL2, and DL3 are defined by the combination of the qualitative relationships of the margins mPS (5) and mPS (6) in the fifth section and the sixth section. Also in this case, the blank portion of the rule table means that the causal relationship between the above-mentioned exhausted NOx margins and the schedule parameters is small or almost nonexistent.

As a part of the emission NOx adjustment rule used in the steam turbine local adjustment (STLT), the minimum emission NOx is set.
In the PS rule table for adjusting the instantaneous value margin and the PA rule table for adjusting the minimum emission NOx average value margin, the high pressure bypass valve operation speed (DAHBV) and the medium pressure bypass valve operation as the three correction schedule parameters are used. Speed (DAIBV), high pressure regulator valve 1st operating speed (DAHC
7 shows a fuzzy rule for correcting V1) in relation to the minimum exhaust NOx instantaneous value margin and the minimum exhaust NOx average value margin in the three target sections shown in FIG. That is, in the PS12 rule table, the minimum emission NOx instantaneous value margin mPS in the first section and the second section
DAHBV, DAIB from the combination of qualitative relationships of (1) and mPS (2)
It defines the correction amount of V and DAHCV1. In the PS23 rule table, the margin mP in the second section and the third section
From the combination of qualitative relationships of S (2) and mPS (3), DAHBV, DAI
It defines the correction amount of BV and DAHCV1. In the PA12 rule table, the minimum discharge N in the first section and the second section
The correction amount of DAHBV, DAIBV, DAHCV1 is defined by the combination of qualitative relations of Ox average value margins mPA (1) and mPA (2). In the PA23 rule table, the correction amounts of DAHBV, DAIBV, and DAHCV1 are defined by the combination of the qualitative relationships of the margins mPA (2) and mPA (3) in the second section and the third section.

Next, the startup schedule correction amount determining means 1
50 will be described. This means comprehensively evaluates the rule-based conclusion 161 (all relevant membership functions and membership values) obtained from the fuzzy rule 160 for each modification schedule parameter to determine the modification amount 151 of the schedule parameter. To do.

As an example, FIG. 11 shows a comprehensive evaluation method of the correction amount KS of the steam turbine-related schedule parameter. In the example of this figure, the membership function and membership value are (NS, 0.6), (Z
O, 0.8), (PS, 0.4), and (PB, 0.2) are shown. The comprehensive evaluation is defined by the weight W (i) of the trapezoidal part determined by each membership value and the center of gravity KSG of the position KS (i). That is, W (1) = 0.126, W (2) =
0.096, W (3) = 0.096, W (4) = 0.09, KS (1)
= -0.15, KS (2) = 0, KS (3) = 0.15, KS (4) =
Since it is 0.35, KSG is calculated as follows.

[0075]

(Equation 4)

Therefore, the correction amount KSG for this activation schedule parameter is 0.0662. This value is passed to the next activation schedule assumption means 110, and a new activation schedule is created by modifying the original schedule parameters according to the following equation.

[0077]

## EQU00005 ## Xi + 1 = Xi + KSG.multidot.Xi --------------------- (5) Here, the schedule parameter Xi is the previous value, Xi + 1
Is a value after correction, i = 0 means an initial value, and a predetermined value is used.

The new activation schedule 111 created by the above method is set again in the high speed prediction means 500.
In the process of repeating such a procedure, the optimality evaluation function 14
At 0, the optimality of the activation schedule is determined each time. Here, the one that satisfies the following judgment conditions is adopted as the optimum startup schedule obtained in the optimum search process.

[0079]

(6) (n> nS). AND. (Min (tSCP (n)) ----------- (6) Here, n is the number of startup searches, and nS is predicted to satisfy all driving restriction conditions in the startup process. It is a preset number of times (for example, nS = 10) and tSCP
(N) is the time required for activation in the n-th activation case, and the one with the shortest activation time is finally adopted as the optimum activation schedule.

FIG. 12 shows the result of prediction of plant dynamic characteristics and the effect of the present invention when the inner optimization loop described above is operated. In this figure, two cases before optimization and one case after optimization are shown as the activation schedule. In the figure, the exhausted NOx shows a moving average value, and the thermal stress is shown as a representative of the rotor surface stress of the high-pressure turbine, but other stresses of interest also show the same tendency. In Case 1, the NOx emissions exceeded the limit value because the gas turbine started up too early, while the latter half of the load increase was late, so the thermal stress satisfied the limit value, but the steam turbine load reached the rated value. Delay, resulting in longer startup time. Contrary to Case 1, in Case 2, the startup of the gas turbine is slow, so the exhausted NOx satisfies the limit value, but on the other hand, the thermal stress exceeds the limit value because the latter half of the load increase is rapid. In addition, the steam turbine load rating is delayed, resulting in a long start-up time. In both cases, even if the exhausted NOx or the thermal stress exceeds the limit value, one of them has a large margin, and an unbalanced and wasteful startup schedule is left. Case 3 is the result of applying the present invention to optimize the startup schedule. Both the exhaust NOx and the thermal stress satisfy the limit values, and the startup can be completed in a shorter time than in the cases 1 and 2. Is shown. However, this conclusion is an inner loop, that is, the high-speed prediction means 500 and the optimum operation schedule search means 100.
Was obtained by using. Although this uses a simplified and accelerated model and evaluation method, even at this stage, it is far more comprehensive, quantitative, and
It is possible to suggest the driving schedule in a form that responds to changes in the situation, and this may be sufficient in some cases.
The present invention has been further devised in order to further improve the application reliability of the obtained operation schedule and not to impair the responsiveness, so as to overcome various obstacles coming from the limit of the calculation function which is actually encountered. To do.

The measure is not to immediately adopt the conclusion in the inner loop, but to evaluate with a more detailed simulator,
In addition, the validity of the conclusion is confirmed in the light of realistic driving restriction conditions, and in the case of non-conformity, the conditions are adjusted again and the optimization search is performed. In other words,
In this method, the inner loop is used to perform repetitive optimization searches, and the outer loop sets the conditions that accurately reflect the reality to improve the final reliability and responsiveness.

A specific configuration example of the outer loop will be shown below.
When the operation schedule 171 at the time of repeating the n-th outer loop is presented, the controller portion 412 replaces this with a control algorithm or setting content control that is as close as possible to the actual machine and moves actuator models such as various valves. , The physical model 411, such as GT, HRSG, and ST, is driven as closely as possible to simulate the thermal-hydraulic behavior of the plant.

The m1 result R1 thus obtained and the m2 result R2 obtained by the high-speed predicting means 500 at the final stage of the repetition generally do not match even though the inputs are the same (see FIG. The difference between (1) a and (2) in 13 corresponds to this). The cause is based on the difference in the fineness of the model and constraints. The status of the difference is the actual machine application determination unit 43.
0 (Fig. 2) is grasped and evaluated. As a result, even if the result R2 is in the operation restriction condition L2, there may be an item in which the result R1 is largely over, and conversely, the result R2 is at the limit and enters the restriction L2, but the result R1 In some cases, there is an item with a margin that is more than the limit L1 (corresponding to (1) b in FIG. 13). Regarding the operation restriction condition L2 of the inner loop and the operation restriction condition L1 of the outer loop, there are many types of evaluation items because the restriction L1 reflects all the actual conditions (m1 ≧ m2). In addition, here, it is assumed that the evaluation items and the results of the respective prediction means have a one-to-one correspondence.
In some cases, it may not be possible.

Therefore, when the comparison is made by the actual machine application determining section 430, the results R1 and R2 have the following two types.

In the first type, the same items are evaluated in the limits L1 and L2, and they can be directly compared with each other. In this example, turbine thermal stress is used (in the case of FIG. 13).

The second type is evaluated in the first predicting means, that is, the high-accuracy simulator, but in the second predicting means, that is, the high-speed simulator, the model is calculated due to the calculation time and the complexity of the model. Omitted, the result R1 is of the species evaluated only by the limit L1, and in this example,
It is assumed that the water levels 306, 307, and 308 in the drum of FIG. 2 correspond to this.

Therefore, the total optimization adjusting unit 420 of FIG. 2 has a function of outputting an instruction 421 divided into the following cases as a boundary condition for the optimization of the inner loop next time.

In case 1A, the result R1 (k) of the k-th (k = 1, m1) evaluation item is the above-mentioned first type and falls within L1 (k) (curve in FIG. 13). (1) b
Equivalent). At this time, δL (k) [= L1 (k) −R1
For (k)], 0 <δL (k) <ε (k) holds, that is, if there is a predetermined appropriate margin value ε (k) and is within the actual limit value L1, Result R1 (k)
From the viewpoint of, there is no request to change the boundary condition of the next inner optimization. When Case 1A occurs for all k, it means that a highly reliable solution applicable to the actual machine is obtained. However, δL (k) ≧ ε (k) ≧
When it is 0, that is, when there is too much margin for the limit value, this effect affects other variables, and therefore a time interval (i ) And the interval before it, the result R
A fuzzy rule for changing the restriction L2 of 2 to a more gradual direction is configured. In this example, the turbine thermal stress limit value SL of the equation (1) is weighted from the first section to the i-th section and is instructed to be relaxed.

In case 1B, the result R1 (k) has the same attributes as in case 1A, but the result is not within the restriction condition L1 (k). In this case, it is possible to add a method to change the method of instructing the inner loop to the fuzzy rule depending on whether or not the time domain that exceeds the limit is the same time period as the result R2 of the high speed simulator. Is. For example, in the situation (a) of FIG. 13, since the curve (1) and the curve (2) have peaks in the same region, it can be seen that the essential difference between the models is small, and the next time (i) Set more stringent constraints on the area. In the situation (b), the curve (1) a is the next time section (i + 1)
Since there is a peak at, the error of the model is considered to be relatively large, the related gain is set in the direction of lower sensitivity, and the limiting condition is moderate in the (i + 1) th interval and The previous section is made strict, and the peak of the next result R2 is driven into the section (i + 1). Further, δL (k) [= L1 (k) −R1 (k)], taking into consideration the change for each number of repetitions (n-1, n-2, ...).
It is also effective to change the boundary conditions for the next inner optimization based on the evaluation of. In the situation (c), there is a peak of the curve (1) a in the section (i-1), so that the instruction is issued in the opposite manner to the situation (b).

In the example of case 1B, contrary to the case 1A, the turbine thermal stress limit value SL of the equation (1) is reset over the region, and all the limiting conditions of the first prediction simulator are satisfied. The model and the limiting condition of the second predicting means are changed in the direction. However, the limiting condition of the first predicting means is unchanged.

Case 2A is the result R of the j-th evaluation item.
1 (j) is the second type described above, and is contained in L1 (j). At this time, 0 <δL (j) <ε (j) holds for δL (j) [= L1 (j) −R1 (j)], that is, a predetermined appropriate margin value ε (j)
And if it is within the actual limit L1, then the result R1
From the viewpoint of (j), there is no request to change the boundary condition for the next inner optimization. However, δL (k) ≧ ε
When (k) ≧ 0, that is, when there is too much margin for the limit value, the indirect method 2A of case 2B described below is instructed for the next optimization.

In case 2B, the result R1 (j) has the same attributes as in case 2A, but is not within the limit L1 (j). At this time, δL (j) [= L1
(J) −R1 (j)] can be evaluated, but there is no evaluation amount that directly reflects this result in the inner loop. In other words, the inner loop does not have the result R1 and the states R2 and L2 corresponding to the constraint L1 as the model and the constraint condition.

Therefore, the next boundary condition of the inner loop is given by the indirect method 2B described below. Indirect method 2B
First, the state of occurrence of δ (j) and its magnitude are grasped for each time interval (i), and the substance relating to other manipulated variables is evaluated from the knowledge of physical phenomena. Speaking of the drum water level, which is a specific example of this example, a situation in which the drum water level R1 (j) suddenly rises in the time section (i) and exceeds the safety upper limit L1 (j) occurs in the nth outer loop simulation. Suppose However, in this case, the simulation in which the plant trips is not performed. At this time, FIG.
Utilizing the knowledge of the physical phenomenon of No. 4, in this example, a valve for increasing the main steam pipe flow rate while making the GT output change rate upper limit value of the second prediction means smaller before the time interval (i) Change the inner loop operation restriction conditions, such as setting the maximum operation speed (CV valve maximum opening speed) to a smaller value. It is possible to change how much the limit value for which operation amount in the causal relationship in FIG. 14 is changed with the same configuration as the measure taken in the inner loop by a method combining a fuzzy relationship map and a rule or neuro. More easily, it can be changed according to the rule "If ...., then ...", and more complex generic algorithms and simple optimal search methods can be found in reference materials such as the Society of Instrument and Control Engineers, "Neuro.・ Fuzzy AI Handbook "
Various methods described in Ohmsha Co., Ltd., Hei 6-5, Part 2 and Part 2 are also applicable. Even in this case, the rule that incorporates the behavior over the time section may be more effective.

Here, as the indirect method 2A, on the basis of the same idea as the indirect method 2B, the reverse operation is instructed to the inner loop, and a gentler operational restriction condition is given to the high speed simulator.

In the formulation for applying the fuzzy theory shown in this example, it is necessary to repeatedly run the simulator on a computer. The reliability of the result is an important factor in applying the obtained optimal operation plan to the actual plant. In order to obtain highly reliable results, it is desirable to use a dynamic characteristics simulator and an operation restriction evaluation function that have characteristics as close as possible to the actual plant. However, if a simulator and an evaluation function that can handle the dynamic characteristics and operation limiting conditions of the entire plant in detail are repeatedly used, it takes a lot of time to obtain an optimum solution even with the current computer. For example, in the problem of optimizing the operation schedule aiming at shortening the start-up time, the plant simulator simulates a phenomenon having a strong nonlinearity from 0% to 100% output of the optimum operation schedule searching means, and the evaluation items are also thermal stresses at each part. Since the exhaust gas characteristics such as NOx and CO2 of the plant vary widely, it may take several tens of minutes to several hours depending on the case. The number of iterations required to converge to the optimal solution varies greatly depending on the optimization technique, formulation, initial assumed state, etc., but considering that it generally requires a dozen to a few tens of times, It is a real obstacle. The present invention provides a structure and means for overcoming this obstacle.

First, in order to speed up the dynamic characteristic simulator used in the second predicting means, the following method is used. (1)
(2) Use a simple model according to the characteristics of the item to be evaluated, in order to reduce the weight of the model, that is, to reduce the overall size of the model by limiting the model to the target plant event. For example, the dynamic characteristics are approximated by static characteristics. (3)
Omitting evaluation of specific state quantities according to phenomena such as start / stop / load change / accident.

In addition, it is desirable to simplify the evaluation of operation limiting conditions such as plant thermal stress and exhaust gas characteristics in order to speed up the process as much as possible. Since the calculation of the thermal stress can take a relatively large time step, a detailed model can be utilized by the second predicting means. In the calculation of exhaust gas, simplification of control operation is introduced.

On the other hand, the simulator used in the first predicting means and the operation restriction condition evaluation function are configured so as to be as close as possible to the characteristics of the actual machine and the operation conditions, and the evaluation of the obtained state quantity is of considerable reliability. Depending on the degree, the level can be applied to the actual machine.

As a result, the speeds of the first and second predicting means can be configured to have a difference of several tens of times or more.

Here, looking at the time required for the computer for this optimization, as shown in FIG. 1, the second prediction is made in response to the command from the central operation room (the place where the power plant is operated). The means determines the dynamic characteristics from the initial time t0 to the end time tf,
Let ΔT2 be the calculation time required to execute once, ΔS2 be the time required to determine the next optimum search direction, and N2 be the number of times until convergence. Similarly, the first predicting means changes from t0 to tf.
The time required for the simulation up to ΔT1, the time ΔS1 required for the optimum operation schedule adjustment, and the number of times until convergence are N1. The time required for the optimization as a whole is given by the following equation.

Tnew = (N2 (ΔT2 + ΔS2) +
(ΔT1 + ΔS1)) N1 On the other hand, if the detailed first prediction means is directly subjected to optimization without taking the hierarchical structure according to the present invention, the required time is given by the following equation.

Tol = N2a (ΔT1 + ΔS2a) Here, ΔS2a: Since it is necessary to search for more parameters, it is larger than ΔS2 (ΔS2a ≧ Δ
S2).

N2a: Since it is necessary to repeat for more parameters, it is larger than N2 (N2a ≧ N
2).

Therefore, the merit of the present invention can be enjoyed when Tnew <Told. In other words,
It is more effective if the speed ratio (= ΔT1 / ΔT2) is sufficiently large and the required number of times N2 required for the optimization adjustment of the outer loop is small.

Normally, ΔT2 tends to increase due to the expansion of the range of equipment handled by the model used, the broadening of the target event, the increase in complexity of the phenomenon, etc. It is desirable to prepare various small-scale ones and use them selectively according to this. On the other hand, since the first predicting means plays a role of verification, it is preferable that the first predicting means accurately represent the actual situation even if it takes some time. It is important to converge the whole by the number of changes, and fuzzy reasoning and neuro are effective methods here as well.

The above example is an example realized mainly by applying the fuzzy theory as an optimization method. However, various other techniques such as the neuron method and the maximum gradient method can be applied. The scope of application of the present invention is not affected by the optimization technique and the formulation of the problem therefor.

In the above-described embodiments of the present invention, the combined cycle power plant will be specifically described, and the thermal stress and the discharge N of the steam turbine and the exhaust heat recovery boiler header will be described.
Although Ox is treated as an operation limiting factor, other factors, for example, the difference between the expansion of the rotor and the casing of the steam turbine, the thermal stress of other parts such as the drum of the exhaust heat recovery boiler, and the emission are determined depending on the characteristics of the applied plant. Examples include SOx and CO2. Also,
It is obvious that even if the thermal stress is not necessarily predicted, the rate of change or the range of change of the steam temperature or the metal temperature may be indirectly controlled as the limit value without changing the essence of the present invention.

Furthermore, in the embodiment of the present invention, the starting schedule parameters to be corrected in the optimization process are set to 8 gas turbine-related and 4 steam turbine-related, but the invention is not necessarily limited to these. However, the present invention can be implemented without changing the basic principle, as long as it is another parameter shown in FIG. 4, for example, a parameter that defines the starting pattern of the plant such as the opening operation timing condition of the regulator valve and the set value for steam pressure control. It is clear that you can do it.

In the embodiment of the present invention, in determining the optimality of the startup schedule, thermal stress and discharge N
I evaluated only the time required to start with Ox as the limiting condition,
Energy loss due to start-up and equipment life consumption are also weighted together with the time required for start-up, and the start-up schedule can be flexibly changed by changing the weighted value depending on the power demand and environmental conditions such as season and time of day. Can be easily realized by applying the present invention.

Further, according to the embodiment of the present invention, the start completion time instructed by the central power feeding command station can be exactly kept and the starting schedule of the shortest time can be created. It is obvious that even if the gas turbine ignition time, the load admission time, the target load arrival time or the like is used instead of the time, the principle of the present invention can be implemented only by shifting the reference time. Furthermore, after the plant actually starts up, if the start completion time is changed by a command from the central power supply command center, or the gas turbine speed is maintained or the load is maintained due to a plant abnormality during startup, and then the error is recovered. According to the present invention, the optimization by rescheduling in the case where the startup is continued by
It can be easily implemented using a dynamic characteristic model.

Operation control system 1 forming the main part of the present invention
As described in the present embodiment, 000 can be used not only for the operation control system of the actual plant but also various other uses. For example, it is an engineering tool for designing a basic plant operation method, various engineering tools for designing a plant, a simulator for operation training, a minor control system design and tuning, or a basic method for supporting a start test. As an engineering tool for designing the basic operation method of a plant, it can be used for reasonably operating methods that consider the dynamic characteristics of the plant and for the formulation of long-term operation plans. As an engineering tool at the time of plant design, it is possible to carry out a limit design concerning the structure, strength, capacity, material, etc. of equipment by simulation in consideration of cost, safety, environmental compatibility, and controllability. Further, when used for operation training, it is possible to learn a quick and appropriate operation method using a simulator in an abnormal situation or an emergency while understanding the relationship between the startup schedule and plant dynamic characteristics. When used for minor control system design and tuning, when executing the optimum operation schedule, properly design minor control systems such as drum level control system and steam temperature control system,
It becomes possible to tune those control parameters in advance.

Furthermore, by expanding the high-precision simulator itself of the first predicting means 410 of FIG. 2 into a system in which the actual plant 300 is replaced, optimization is performed in such a manner that repeated start-up tests at the site are learned one after another. It is possible to significantly reduce the number of adjustment work steps.

Further, in the embodiment of the present invention, the thermal power plant operation control system is a combined cycle mainly composed of an ordinary thermal power plant consisting of a boiler and a steam turbine, a gas turbine, an exhaust heat recovery boiler and a steam turbine. The application to a power plant was explained, but a gasification combined cycle power plant having a coal gasification furnace for producing fuel for a gas turbine, and a boiler that burns coal in a fluidized bed instead of an ordinary burner are described. Even in the normal pressure or pressurized fluidized bed boiler power plant used, the basic principle of the present invention can be easily implemented without changing.

The present invention can be applied to other power plants, for example,
It is needless to say that the present invention can be applied to a normal power plant including a boiler, a steam turbine and a power generator, a coal gasification power plant, an atmospheric pressure or a pressurized fluidized bed boiler power plant. It is also clear that the fuel used is not limited to coal, petroleum, LNG and the like.

Further, not only for thermal power generation but also for more complicated nuclear power plants, reference literatures, for example, Akiyama "Atomic Plant Engineering", Corona Publishing Co., Sho 54,
The present invention can be applied in exactly the same manner by appropriately giving the nuclear model as described in Chapter 3 and the accompanying limiting conditions.

For example, in a boiling water nuclear power plant, a reactor pressure vessel and a containment vessel, a steam turbine for converting generated steam into rotational energy, a condenser for condensing steam, and a feed water for returning condensed water to the reactor. System, a recirculation system for controlling generator output, etc., and as operation restriction conditions, at least the nuclear reaction period, control rod movement speed, nuclear reaction output change speed, cladding of all fuel rods. Resistance limit value, hydrogen concentration in reactor water, oxygen concentration in reactor water, recirculation pump flow rate change rate, reactor pressure and rate of change, reactor pressure vessel temperature rate of change, purification system flow rate and temperature, main shutoff valve Front-to-back differential pressure and temperature difference, steam turbine rotor internal and external thermal stress, and moisture separator internal water level change range, main steam pipe thermal stress, reactor water level change range, maximum feed water flow rate, main steam system control valve and stop valve Opening and closing speed , The main steam maximum pressure value, the maximum flow rate in the feed water heater pipe, and the condenser temperature / waste water temperature variation range, all or part of which falls within the respective specified ranges, the start-up time is shortened and high economic operation method This is useful for creating a plant operation plan for the purpose of achievement.

Further, in the case of a pressurized water nuclear power plant, it is composed of a reactor pressure vessel and a containment vessel, a steam generator, a steam turbine, a condenser, a water supply system for returning condensed water to the steam generator, a generator and the like. As operation restriction conditions, at least the nuclear reaction period, control rod movement speed, boric acid water injection flow rate, nuclear reaction output change rate, all fuel rod cladding tube resistance limit value, reactor water hydrogen generation concentration, reactor Water oxygen concentration, primary coolant pump flow rate change rate, reactor and steam generator pressure and change rate, reactor pressure vessel temperature change rate, purification system flow rate and temperature, main block valve differential pressure and temperature difference , Thermal stress inside / outside of steam turbine rotor, water level change in feed water heater, primary low temperature side and high temperature side pipe thermal stress, main steam pipe thermal stress, steam generator water level change range, maximum feed water flow rate, main steam system control valve And stop valve opening / closing speed, main steam All or part of the maximum pressure value, the maximum flow velocity in the feed water heater pipe, and the temperature variation range of the condenser hot drainage temperature are kept within their respective specified ranges while aiming to shorten the startup time and achieve high economic operation. It can also be expanded as a plant operation plan creation device.

As described above, according to the present invention, it is too complicated to approach from the front such as a power generation system,
Two types of high-precision and high-speed simulators are placed on the outside (upper side) and inside (lower side) of the inner loop when optimizing operation plans, control methods, and emergency measures for large-scale systems. Optimization and outer-loop optimization can be achieved by selecting an appropriate optimization method and load distribution.

[0119]

The first effect of the present invention is that, in the operation control system of the power plant, the equipment life which was difficult by the conventional method under the practical capacity limit of the computer and the application limit of the optimization technique. Automatic creation and execution of an optimal operation schedule that minimizes start-up time, minimizes stop time, responds to load changes, and stabilizes accident measures while simultaneously satisfying operational restrictions and environmental regulations such as exhaust NOx Can be realized while considering the limits of realistic computer capabilities. As a result, the burden on the operation planner and the operator can be significantly reduced, and the energy loss accompanying the shortening of the start-up time can be reduced, so that the operation cost of the power plant can be significantly reduced.

The second effect of the present invention is to minimize start-up time and stop-time while simultaneously satisfying operational restriction conditions such as equipment life and exhaust NOx and environmental regulation values in the start-up test process of the power plant. It is possible to speed up the on-site adjustment of the optimal operation schedule for quick response to load changes and stabilization of accident measures. As a result, the burden on the operator and the designer is significantly reduced, and the energy loss due to the shortened adjustment during the start test can be reduced.

The third effect of the present invention is that, in the operation control system of the power plant, while satisfying operation restriction conditions such as equipment life and NOx emissions and environmental regulation values, at the same time specified by the central power supply command center. It is possible to complete start and stop. As a result, stable and accurate power supply to the power system, which requires frequent start and stop of the power plant due to fluctuations in power demand, becomes possible.

The fourth effect of the present invention is that in the operation control system of the power generation plant, the operation limit value is changed during the start-up of the plant, the designated start / stop completion time is changed from the central power supply command station, or the abnormality occurs. Even if re-scheduling is required after restoration or in an emergency, it is possible to automatically generate an optimum start / stop schedule and execute it. This enables flexible and safe plant operation and power system operation.

[Brief description of the drawings]

FIG. 1 is a process flow diagram illustrating a basic configuration of the present invention.

FIG. 2 is a block diagram showing a basic configuration and a device configuration of a start control system for a combined cycle power plant that is an embodiment of the present invention.

FIG. 3 is a conceptual diagram showing a basic idea of an inner loop activation schedule optimizing means of the present invention.

FIG. 4 is an explanatory diagram (inner loop) showing a relationship between a plant starting process and a starting schedule parameter.

FIG. 5 is a graph showing a turbine thermal stress characteristic evaluation method.

FIG. 6 is a graph showing an exhaust NOx characteristic evaluation method.

FIG. 7 is an explanatory diagram (inner loop) showing a relationship between a fuzzy rule for optimization of a startup schedule and a modification schedule parameter.

FIG. 8 is an explanatory diagram (inner loop) showing a membership function used in a fuzzy rule.

FIG. 9 is an explanatory diagram (inner loop) showing the overall configuration of a fuzzy rule.

FIG. 10 shows a part of a turbine thermal stress adjustment rule used in the steam turbine master plan (STPS) (inner loop).
FIG.

FIG. 11 is an explanatory diagram showing a method for determining a startup schedule correction amount.

FIG. 12 is an explanatory diagram showing an operation result of the inner loop of the combined cycle power plant start control system to which the present invention is applied and its effect.

FIG. 13 is an explanatory diagram for comparing the results of the first prediction means (outer loop) and the second prediction means (inner loop) and the limiting conditions.

FIG. 14 is an explanatory diagram showing an example of knowledge about a physical phenomenon for instructing a correction amount to a second predicting unit (inner loop).

[Explanation of symbols]

10 ... Operation planner, 100 ... Optimal operation schedule search means, 120 ... Operating characteristic evaluation function, 122 ... Turbine thermal stress characteristic evaluation function, 123 ... Boiler thermal stress characteristic evaluation function, 124 ... Emission NOx characteristic evaluation function, 125 ... Startup required time evaluation function, 140 ... Optimality evaluation function, 150 ... Startup schedule correction means, 160 ... Fuzzy rule, 1
70 ... Startup schedule setting means, 180 ... Startup schedule display means, 200 ... Device control system, 300 ...
Combined cycle power plant, 310 ... Exhaust heat recovery boiler equipment, 320 ... Gas turbine equipment, 330 ... Steam turbine equipment, 400 ... Optimal operation application adjusting means, 410 ... Optimal operation application adjusting means, 411 ... Physical model, 412 ... Controller part , 420 ... Overall optimization adjusting unit, 430 ... Actual machine application determining unit, 500 ... High speed prediction means, 532 ... Dynamic characteristic model, 533 ... Bus turbine model, 534 ... Boiler model, 535 ... Steam turbine model, ... 1000 ... Operation control system.

Claims (11)

[Claims] 1. A device for planning and presenting an operation plan of a plant, comprising a dynamic characteristic simulator having a high prediction accuracy of a plant state with respect to an operation setting, and a function for evaluating in detail operation limiting conditions of an actual machine. A predicting means, a dynamic characteristic simulator that predicts a plant state faster with respect to an operation setting, a second predicting means having a function of evaluating an operation restriction condition at a higher speed, and the second predicting means. Optimum search means for automatically searching the optimum operation setting contents of plant operation with respect to at least one event of start-up, stoppage, load change, and accident, and the optimum operation setting obtained by the optimum search means The difference between the result obtained by inputting the content into the first predicting means and the result of the second predicting means in the optimum operation setting content is evaluated, and the result is compared with the second predicting means. Then, the operator is instructed to modify the operation restriction condition and / or the model of the dynamic characteristic simulator and the input constant, and / or to change the operation setting content, and cause the optimum search means to perform the optimization again. A plant operation plan preparation device comprising: an optimum operation application adjusting means having a function of repeating a series of procedures until the result of the first predicting means satisfies an operation restriction condition of an actual machine. 2. The apparatus according to claim 1, wherein the second
The dynamic characteristic simulator used in the prediction means and the operation restriction condition evaluation function are prepared only in types according to the event to be optimized and the goal of the optimization, and either of them is selectively used. Plant operation planning device. 3. The apparatus according to claim 1 or 2, wherein
A plant operation plan creating apparatus characterized in that a plurality of the optimum search means are prepared and any one of them is selectively used in accordance with an event to be optimized and / or a goal of the optimization. 4. The apparatus according to claim 1, 2 or 3, wherein a plurality of the optimum operation application adjusting means are prepared according to the event to be optimized and / or the goal of the optimization. A plant operation plan creation device characterized by selectively using one of them. 5. The apparatus according to claim 1, wherein the plant is at least a gas turbine, a boiler that generates steam by exhaust gas heat of the gas turbine, and the generated steam is converted into rotational energy. It consists of a steam turbine, a condenser for returning steam to the boiler to return it to the boiler, and a generator. As the operation limiting conditions, steam turbine thermal stress and plant exhaust gas components, boiler piping thermal stress, and fuel supply to the gas turbine are provided. Volume change rate, gas turbine rotation speed change rate, gas turbine purge operation time, gas turbine rotation speed holding level and holding time, gas turbine load change rate, steam drum water level fluctuation range, maximum feed water flow rate, main steam system control valve and Stop valve opening and closing speed, main steam maximum pressure value,
A plant operation planning apparatus, characterized in that at least a part of the main steam temperature variation range and the condenser hot drainage temperature variation range are set within respective prescribed ranges. 6. The apparatus according to any one of claims 1 to 4, wherein the plant is at least a gas turbine, a furnace for supplying gas produced by combustion of coal to the gas turbine as fuel, and exhaust gas from the gas turbine. A boiler that generates steam by gas heat, a steam turbine that converts the generated steam into rotational energy, a condenser that returns the steam to the boiler,
And a generator, and as the operation limiting condition,
Range and rate of change of volumetric flow rate and temperature of furnace generated gas,
Thermal stress inside and outside the rotor of steam turbine and plant exhaust gas component, gas turbine rotation speed change rate, gas turbine purge operation time, gas turbine rotation speed holding level and holding time, gas turbine load change rate, boiler pipe thermal stress steam drum water level fluctuation Width, maximum feed water flow rate, main steam system control valve and stop valve opening / closing speed, main steam maximum pressure value, main steam temperature change width, condenser temperature waste water temperature change width, at least part of each within the specified range. A plant operation plan creation device characterized by being able to enter. 7. The apparatus according to any one of claims 1 to 4, wherein the plant is at least a boiling water reactor and a containment vessel, a steam turbine for converting generated steam into rotational energy, and steam recovery. It consists of a water condenser, a water supply system for returning the condensate to the reactor, and a generator.The operation limiting conditions include the nuclear reaction period, control rod movement speed, nuclear reaction output change speed, and total fuel rod Cladding tube resistance limit value, Reactor water hydrogen generation concentration, Reactor water oxygen concentration, Recirculation pump flow rate change rate, Reactor pressure and change rate, Reactor pressure vessel temperature change rate, Purification system flow rate and temperature, Main blockage Differential pressure across stop valve and temperature difference,
Thermal stress inside and outside the rotor of steam turbine, water level change range in moisture separator, main steam pipe thermal stress, reactor water level change range, maximum feed water flow rate, main steam system control valve and stop valve opening / closing speed, main steam maximum pressure value A plant operation planning apparatus, wherein at least a part of the maximum flow velocity in the feed water heater pipe and the temperature variation range of the condenser hot water drainage temperature is within the respective specified ranges. 8. The apparatus according to claim 1, wherein the plant is at least a pressurized water reactor pressure vessel, a containment vessel, a steam generator, a steam turbine, a condenser, and a condensate. A water supply system for returning the steam generator to a steam generator, and a generator, and as the operation limiting condition, a nuclear reaction period,
Control rod movement speed, boric acid water inflow / outflow rate, nuclear reaction output change rate, cladding fuel resistance limit of all fuel rods, reactor water hydrogen generation concentration, reactor water oxygen concentration, primary coolant pump flow rate change rate,
Pressure and rate of change of reactor and steam generator, temperature change rate of reactor pressure vessel, flow rate and temperature of purification system, differential pressure and temperature difference between main block valve, steam turbine rotor internal and external thermal stress, feed water heater Inner water level change width, primary low temperature side and high temperature side piping thermal stress,
Main steam piping thermal stress, steam generator water level change width, maximum feed water flow rate, main steam system control valve and stop valve opening / closing speed, main steam maximum pressure value, feed water heater maximum pipe flow velocity, condenser hot drainage temperature change A plant operation plan creation device, characterized in that at least a part of the width is within each specified range. 9. The apparatus according to claim 1, wherein an actual plant is used instead of the simulator used in the first predicting means, and a state obtained from the actual plant is also displayed. A plant operation plan creation device configured to evaluate the operation restriction conditions. 10. The apparatus according to any one of claims 1 to 9, wherein the second prediction means is an operation schedule assumption means for predicting an operation schedule of the plant in advance, and the plant according to the operation schedule. A high-speed predicting means for predicting a driving characteristic on the assumption that the vehicle has been driven, a driving characteristic evaluating means for evaluating a driving characteristic obtained by the high-speed predicting means by comparing it with a driving restriction condition, and a driving characteristic evaluating means for obtaining the driving characteristic. Based on the operation characteristic evaluation result, the operation schedule correction means for correcting the assumed operation schedule to obtain better operation characteristics, the operation schedule assumption, plant operation characteristic prediction, operation characteristic evaluation, operation An optimality judgment method for judging whether or not the operation schedule has converged to the optimum value in the process of repeating the schedule modification. And an operation schedule setting means for setting the obtained optimum operation schedule in the device control system as the optimum operation setting content, and an operation schedule display means for presenting the operation schedule to an operator via a display device. A plant operation plan creation device comprising: 11. A first predicting means having a dynamic characteristic simulator having a high prediction accuracy of a plant state with respect to an operation setting, and a function of evaluating an operating restriction condition of an actual machine in detail,
A method for planning and presenting an operation plan of a plant by using a second predicting means having a dynamic characteristic simulator having a higher prediction speed of a plant state with respect to an operation setting and a function of evaluating an operation restriction condition at a higher speed. And (a
1) At least one event at the time of startup, shutdown, load change, and accident is input to the second predicting means with an initially assumed operation schedule according to initial conditions to predict the dynamic characteristics of the plant, (A2) Evaluating the predicted dynamic characteristics of the plant in light of operation limiting conditions, (a3) If the evaluation result does not satisfy the operation limiting conditions, correct the operation schedule, and proceed to step (a1). Returning to (a4), when the evaluation result satisfies the operation limiting condition, the operation schedule at that time is set as the optimum operation setting content of the plant operation, and the process proceeds to the next step (b1), and (b1) the optimum operation setting The contents are input to the first predicting means to predict the dynamic characteristics of the plant, (b2) the predicted dynamic characteristics of the plant are evaluated in the light of the operation restriction conditions of the actual machine, and (b3) the evaluation result is Driving system When it is determined that the condition is not satisfied, the difference is small based on the difference between the evaluation result obtained in step (b2) and the evaluation result obtained in step (a2) for the optimum operation setting content. In the same direction as above, the operation limiting condition relating to the second predicting means is corrected, and / or the model of the dynamic characteristic simulator and the input constant are corrected, and / or the operation setting content is corrected.
Returning to 1), when it is determined that the evaluation result satisfies the operation limiting condition of the actual machine in (b4) step (b2), the optimum operation setting content is set as the final operation setting content of the actual machine. How to create a plant operation plan.