JPS6232388A - Automatic driving device for control rod - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention]

[Field of Application of the Invention] The present invention relates to an automatic control rod operation device for an atomic couplant 1-, and in particular, to automatically start up an atomic couplant in a short time while complying with restrictions on reactor operation. This invention relates to an automatic control rod operating device suitable for [Background of the Invention] In the boiling water nuclear power generation plant 1-, approximately 200 control rods are withdrawn and inserted in order to control the reactor output during plant startup, shutdown, control rod pattern replacement, etc. An operation is being performed. Conventionally, when operating control rods, operators would use their years of experience and knowledge to understand the state of the plant based on information observed in the plant, and then manually operate the control rods in response to that state. is operating. however,
Recently, automation of control rod operations has been desired for purposes such as shortening plant start-up time, saving labor in operation, and preventing erroneous operations. Bran 1 - Briefly explain how control rods are operated during startup, divided into (1) reactor critical operation, (2) reactor critical operation, and (3) reactor power increase operation. . The critical operation of the nuclear reactor (1) is an operation in which the reactor is brought into a critical state while sequentially withdrawing the control rods when the reactor output is less than 1%. Normally, in order to suppress a sudden increase in reactor output, control rods are operated so that the reactor period does not become less than 50 seconds, for example. If reactor power suddenly increases, a separate interlock is activated and the reactor is automatically brought to a scram stop. After completing the reactor temperature and pressure raising operation ci in (2), and the operation in (1), the control rods are withdrawn one after another to determine the reactor temperature (reactor water temperature).
from about 100°C to about 265°C at a pressure of 1 kHz/cm.
The temperature and pressure of the reactor are increased at a rate of about 5°C/h or less in order to prevent thermal shock to the reactor structural materials. The reactor power increase operation (3) is an operation in which the reactor power is increased at a rate of, for example, 300 M W e / h or less by operating the control rods after the reactor temperature and pressure increase operation (2) is completed. As a prior art related to automatic operation of control rods for the purpose of automating such operation, Japanese Patent Application Laid-Open No. 55-114990 ``Reactor Power Control Device for Nuclear Power Plant'' is known. This technology states that the control rods to be operated and the amount of control rod operation are determined in accordance with the reactor output set value and load fluctuations output from other devices. However, when this technique is applied to the automatic operation of control rods in (1), (2), and (3) above, there are the following disadvantages. (a) It is stated that the control rod operation i is determined according to the reactor output set value or load fluctuation, but the method is not disclosed.The value of the control rod is that the control rod to be operated is It changes depending on whether the control rod is pulled out from the top or bottom of the core, whether the control rods surrounding the control rod are pulled out from the top or bottom of the core, etc. Therefore, in the conventional technology,
Example λ-ba. If the reactivity applied to the control rods is too high, the change in reactor power may become too large, or if the reactivity applied to the control rods is too small, the change in reactor power may be too small. (b) The above-mentioned prior art does not include any description regarding when to operate the control rods. Therefore, by operating the control rods one after another,
There is a uJ capability that causes sudden changes in the reactor output. (C) Since the control rods are not configured to be operated by inputting reactor temperature, reactor period, etc., if applied during reactor critical operation, reactor temperature rise, and pressure rise, operational limitations may occur. 5. For example, reactor power may increase rapidly and bring the reactor to a scram shutdown. If the reactor is brought to a scram stop in this way, or conversely, the rate of increase in the reactor's output is too low, both

[Since it does not have F control, it naturally has the disadvantage that the 1111 force layer becomes much J when started. As described above, the conventional technology has the drawback of not being able to provide an automatic control rod operating device that can start up the reactor in a short time while complying with operator restriction conditions such as reactor temperature change rate and reactor period. be. [Object of the Invention] The object of the present invention is to comply with the control matters for ``atomic P operation'' and to achieve a short time i1! It provides an automatic control system that can automatically start the reactor. “Summary of the Invention” The present invention aims to reduce operations conventionally performed by operators to the greatest extent possible.
In order to automate the @star, it is necessary to take the same way as the operator's way of thinking and automate the 8Fs. Second, we focused on the fact that in order to realize the A response method of operation Ei, it is necessary to have a means to learn how to operate in the same way as the operator.
One of the features of the present invention, which was born from the
The amount of operation of the control rod, the coordinates of the control rod (coordinates that distinguish whether the control rod is near the center or the periphery of the reactor core), and the usage position of the control rod (whether the tip of the control rod is at the top or bottom of the core) the withdrawal position t!1), the withdrawal position of the control rod around the control rod, the reactor output (or what distinguishes whether it is a reactor critical operation, a temperature and pressure increase operation, or a reactor power increase operation) Alternatively, it can be determined based on the thermal margin of the reactor core (margin with respect to the limit value of the limit power ratio or linear power density). In addition, since it is necessary to determine the amount of operation from various factors as described above, the fish is also characterized by determining the number of operations using fuzzy logic. Note that the control rod to be operated is determined from the control rod operation sequence. Control rod operation timing is determined by the reactor section in order to comply with operational restrictions, and during li'i' operation, the reactor period is greater than or equal to the target value (for example, 200 seconds) or is negative, and the reactor temperature is increased. During boosting operation, the reactor temperature (reactor water temperature or reactor structural material temperature) is lower than the target reactor temperature (changes over time), and when increasing reactor power, the reactor output is lower than the target output (time-varying). Even if the condition that it is lower (changes in terms of In this way, during the reactor startup process, the 7 & criteria used to determine the timing of operation change, so it is important to be able to respond flexibly to this. The operation timing is determined by a production system using the 1n-(:hsn rule.Another feature of the present invention is that in order to realize the operator's adaptive operation method, the control rod integrated controller is equipped with a control rod operation amount control system. A means for learning the determination method is provided. 1 At a certain point 7 In other words, after operating the control rods, changes in reactor power, changes in neutron flux,
It measures the time interval between control rod operations, and for example, if the change in reactor output is excessive, it can be used to ensure that the change in reactor output is appropriate the next time the same operation is performed. Some of the data used to determine the control rod operating amount will be corrected. For example, if the control rods are operated too frequently, some of the data used to determine when to operate the control rods is similarly corrected so that the time intervals between the operations are appropriate. [Embodiment of the Invention] An embodiment of the present invention will be described below with reference to FIG. In the figure, 1 is a boiling water reactor, 2 is a control rod, 3 is a reactor core, 4 is a recirculation pump, 5 is a main steam pipe, and 6 is a water supply pipe. 10 is a control rod control device consisting of a control rod driving device 11 and a control rod dynamic control device 12; 20 is a control rod operation timing determining section 21, a control rod operation amount determining section 22, and a control rod operation amount learning section 23; This is a control rod integrated control device consisting of: a is the main steam flow rate, b is the neutron flux, and C is the control rod pattern signal and control rod operation end signal. d is operated 1
This is a signal indicating the coordinates of one or more control rods (for example, #(22-43)), the amount of operation thereof (for example, 4 notches), and an instruction to start the operation. W is a control rod drive signal or a signal indicating a control rod pattern. 6e is a signal related to the thermal margin of the reactor core 3 calculated by another device (such as the limit power ratio or linear power density), and f is the output from another device. set value for the main steam flow rate that changes over time (hereinafter referred to as the reactor power set value)
1g is determined based on the control rod operation sequence, the control rod coordinates to be operated, the control rod pattern, and the main steam flow J
ia (hereinafter also referred to as reactor power), thermal margin of the reactor core 0
This is a signal containing h is a calculation instruction signal to calculate the manipulated amount Δ2 of the control rod to be operated from now on, p is the calculated manipulated amount Δ2 of the control rod (
For example, ΔZ=4 notches). q is a calculation instruction signal to calculate the state change of the plant since control rod operation has started, and r is a calculation instruction signal to start learning calculation for modifying the control rod operation amount determination rule. S is a signal including a power deviation, which is a deviation between the reactor power set value f and the reactor power a, and a neutron flux. t is information regarding the rule used to determine the manipulated variable, and U is information for modifying the rule used to determine the manipulated variable. Hereinafter, the operation in this embodiment will be described by taking as an example a reactor power increasing operation in which the reactor power is increased while automatically operating the control rods. Note that during reactor critical operation and reactor temperature and pressure increase operations, the basic operations are the same, with the only difference being the types of parameters used. The control rod operation timing determination unit 21 of the control rod integrated control bag M20 determines whether or not to operate the control rods based on the deviation between the reactor output a and its target value f, and determines whether or not to operate the control rods. When it is determined that the control rod operation amount is good, the control rod operation amount learning section 23 performs a learning process as described later, and then outputs the signal to the control rod operation amount calculation section 22. When the control rod operation amount calculation unit 22 receives a calculation start instruction signal. Input the signal g including the coordinates of the control rod to be operated determined based on the control rod operation sequence, the control rod pattern, the reactor output a, and the thermal margin e of the reactor core, calculate the control rod operation amount ΔZ, and calculate the control rod operation amount ΔZ. The resulting operation amount ΔZ is output as a signal P to the control rod operation timing determining section 21. After calculating the operation amount,
The control rod operation timing determination unit 21 determines the coordinates of the control rod to be operated, its operation amount (value of ΔZ), and operation start instruction d determined based on the control rod operation sequence. ! Output to 12. The control rod drive controller 12 drives the control rods 3 through the control rod drive bag fill according to the instruction d.
Pull it out. On the other hand, when the instruction d to start control rod operation is output, the control rod operation timing determining section 21 outputs an instruction q for calculating a change in the state of the plant to the control rod operation amount learning section 23 . When the control rod operation amount learning unit 23 receives the calculation instruction q,
Signals such as output deviation. While inputting the values, calculations of the positive and negative maximum values of the deviation of the reactor output, the maximum value of the neutron flux change rate, the time interval of control rod operation, etc. are started. When the control rod operation end signal C is output to the control rod operation timing determination section 21, the initial state is returned. and. The control rod operation timing determination unit 21 determines whether the control rods should be operated again based on the output deviation and the like. When it is determined that the control rod is to be operated again, the control rod operation timing determining unit 21 outputs an instruction r to start learning to the control rod operation amount learning unit 23. At this time, the learning unit 23 completes the calculation of the plant state change and determines whether the previous control rod operation amount was appropriate based on the calculated output deviation, the maximum value of the neutron flux change rate, and the time interval of control rod operation. An evaluation is made as to whether or not. If it is determined that it is inappropriate, the rule used to calculate the operated vehicle ΔZ is corrected. Thereafter, the operation amount ΔZ is calculated in the same manner as described above based on the revised rule, and when the calculation is completed, the control rod withdrawal operation is started again based on the instruction from the control n operation timing determining section 21. As described above, in this embodiment, the control rods are automatically operated while the control rod integrated controller [20] determines when to operate the control rods, learns, and determines the amount of operation. Next, the configuration of the control rod integrated control device 20 for realizing all the functions as described above will be explained in detail using FIG. 2. The control rod operation timing determining section 21 includes a data input section 211, an input data file 212. Occurrence event monitoring/registration unit 213
．． A file 214 that stores past data and events corresponding to the data. Knowledge base 215. It is composed of an interpreter (inference mechanism) 216, an operation instruction output section 217, and the like. The knowledge base 215 stores the following first-order rules. Rule #42 regarding control rod withdrawal operation Rule #43 Rule #43 monitoring output deviation The event monitoring/registration unit 213 periodically repeats the process shown in FIG. While comparing with the file, events that have changed (occurred) (for example, control rod operation is completed) are stored in the event table of the knowledge base. FIG. 4 shows the configuration and operation of the knowledge base and ink printer. In the interpreter, the following series of operations are performed. (1) Capture the events that have changed (occurred) from the event table,
The condition of the rule related to the event is found by passing it through the discrimination net, and the condition part of the rule related to the changed event is activated. (2) If the status section is active, a rule for which all conditions are active is given a priority and registered in the agenda list, which is a place where executable rules are stored. (3) Select one of the rules with the highest priority from among the rules registered in the assignee list and execute the operation section0
When the above processing is completed, the process returns to the beginning, and if the event has changed, processes (1) to (3) are performed, and if the event has not changed, process (3) is performed. In such processing, when the rule #43 is selected, it is determined whether the output deviation is greater than or equal to a reference value. That is, since the output deviation Y becomes a variable, the value of this variable is extracted from the input data file 212 and it is determined whether it is 1% or more. On the other hand, when rule #42 is selected, the coordinate X of the control rod to be operated is calculated after the control rod operation amount determination unit 23 has learned. That is, from the control rod operation sequence and control rod pattern, the coordinates of the next control rod to be operated are determined, and the instruction (signal R) to calculate the operation amount ΔZ is as follows:
This is output to the control rod manipulated variable determination section 22, and the operating section ΔZ is calculated using the method described below. After that, an instruction to start control rod operation (control rod seat 11x, operation amount ΔZt & signal (1) including signal (1) is outputted via the operation instruction output unit 2J-7. Next, a method for calculating the control rod operation number ΔZ The control rod operation amount determination unit 22 determines the rules used to determine the operation,
It consists of a membership function, a file 221 that stores the membership value of the rule used last time, and a section 222 that calculates ΔZ. The rule used to determine the amount of operation is a rule that describes the qualitative knowledge of transfer i1 as follows. Rule #I PM)) Here, PI3 is InositivQRig+ P S
is P ositivs S wall + P M is Po
Taste gitiveMedia&. Using such rules, fuzzy reasoning (vague logic) is used when determining the operator. The reasons for this are: (i) Since the amount of operation is determined using qualitative reasons with operation pages, there is no need for complex algorithms;
i) It is easy to calculate even if there are many variables. The fuzzy inference method will be simply described using the simplified rule shown in FIG. Here, the membership value μ indicates the ae to be exhausted. That is, if the extraction positions are from 0 to 48, P8 means that the position is 18 < leprosy, so if the position is exactly ]8, μ is 1.0.
If 24, μ is 0, if 5.30, μ is O, the fifth
The rule in the figure is a rule in which only two conditions of the rule marked L are taken out and simplified for the purpose of explanation. The coordinates of the control rod to be operated and its extraction position can be determined from the control rod operation sequence and control rod butterfly, so the current r, Z
The degree to which the condition 9 rule #i where rQ and Zo are determined μml is satisfied, and the degree to which condition 2 is satisfied is 11r2.
etc., and set its minimum value μ as the membership value of rule #j. The operation amount is ΔZ and is 7,000 l1. Since a plurality of such rules are available, the weighted average value of the degree μ to which each rule is satisfied f- and the operation amount ΔZ2 is taken as the operator ΔZ of the control rod who actually operates the control rod. FIG. 6 shows the calculation procedure. Among these, the reason why μ虞 is stored is to remember the rules used.
A rule with a large value of p is a rule that greatly contributes to determining the manipulated variable. Next, the control rod operator learning section 23 will be described. The control rod operator learning section 273 includes a file 231 storing rules for learning and their membership functions, and a calculation section 232. After control rod operation, change in plant status (maximum value on positive side of output deviation EI's + axt
Maximum value Er□ on the negative side. , neutron flux change rate φ. maximum value φ
c1), and the procedure for calculating the time interval of control rod operations,
It is shown in FIG. Examples of rules for learning are as follows. Rule #k IF' (maximum value on the negative side of output deviation E rptn is medium) THEN (one operator determination rule is modified so that the amount of operation is small) Rule # (k+1) IF (Neutron flux Maximum value of rate of change φ., 1 is large) THEN
(Modify two rules for determining the amount of operation so that the operation value becomes smaller) Rule # (k + 2) II? (mj interval T is too short during control rod operation) THEN
(One operation amount determination rule is modified so that the operation panel becomes larger) Whether or not these learning rules should be applied is similarly determined based on fuzzy inference. That is, after the constraint a1 operation, the membership value μ, which is the degree to which the learning rule is satisfied, is calculated for each learning rule, and the maximum value μk of μ is calculated.
If MI&X is, for example, 0.8 or more, the manipulated variable determination rule is actually modified. Here, for example, to make the amount of operation smaller, just one rule (8) means to correct the rule that actually contributed the most (μ is the largest) among the rules for determining the amount of operation.The amount of operation is small. 4 Correct it so that Δ
If Z is PM, modify 8Z to become PS. When modifying two rules, among the operation amount determination rules, the rule with the largest μ and the rule with the first largest μ are modified. These calculation procedures are summarized in FIG. Next, the results of a simulation test of the embodiment of the present invention will be described. In the simulation test, in order to specifically confirm the effectiveness of learning, tests were conducted separately for cases without the learning function (control rod operation amount learning section 23) and cases with the learning function. First, Figure 9 shows the results when there is no learning function. This will be explained using FIG. 0 and FIG. 1. Furthermore, it is assumed that the event of rule #42 "Break point is output increase H, M control rod operation ends" has already occurred at the start of the simulation.6 Figure 9 shows the flowchart for determining the manipulated variable. This is the result if the rules were appropriate from the beginning. When the deviation (output deviation or load deviation) between the reactor output Power and its set value (set load) exceeds the reference value, the rule #42 is selected (4), and the manipulated variable ΔZ is determined by fuzzy reasoning (1). , control rod withdrawal started and Maya 5
．． Note that here, the manipulated variable ΔZ corresponds to the application of control rod reactivity. At this time, the control rod reaction gradually increases, the neutron flux increases, and the nuclear/reactor power (load) follows the set load and increases by 1. Thereafter, when the control 3i1 frame is Hiiragi-rL again and the load deviation exceeds the base wall value, the same operation as above is performed. The test results show that if the manipulated variable ΔZ is appropriate, the control rod can be operated automatically (f:ly is 1-') and the load can be automatically raised following the set load. Figure 10 shows the result "r!" when the manipulated variable ΔZ is initially excessive in the rule for determining the manipulated number ΔZ. Since the manipulated variable ΔZ is excessive, the deviation from the set load is very large. Preferable for more stable operation of the nuclear reactor 1.<
On the other hand, Fig. 11 shows the manipulated variable Δ
This is the result when Z is too small. In this case, it can be seen that even if the control rod is pulled out frequently, the operation that follows the set load cannot be achieved and the startup time becomes longer.7
Note that in the conventional general method of predetermining the manipulated variable by off-line calculation, it is necessary to estimate the manipulated variable ΔZ on the safe side (smaller), so it is thought that a similar result will be obtained. As a result of J-Metama, in order to more stably increase the reactor output while shortening the startup time, it was found that a learning function that automatically corrects the rules that determine the manipulated variable ΔZ is necessary. I see.Next, the simulation results when the learning function (control rod operation learning unit 23) is added will be described using FIGS. 12 and 13. FIG. 12 shows the result when the manipulated variable ΔZ was initially made too large (corresponding to the manipulated variable in FIG. 10). At this time, at first, the learning rule #k was applied because the maximum value on the negative side of the load deviation was medium. the result,
When the control rod is operated again, this load deviation will be approximately 1/2
It was found that the reactor power can be stably increased by reducing the power to 100%, while the case in FIG. 13 is similar to that in FIG. 11. This is a characteristic when the initial operation amount is too small. In this case, since the time interval between control rod operations is very short, the learning rule # (k + 2) is applied twice, and the operation amount is Determined! As a result, the control rod operation preference is corrected compared to Fig. 11, and the set load number J automatically follows and the no. ↓It has been confirmed that operation with reduced operating time can be achieved. The effect of applying 6 fuzzy inferences to the learning process for determining the control rod operating area is that it is possible to secure control performance equivalent to or better than that of the operator using a simple algorithm that uses many variables. -The development and maintenance of software is extremely important. According to this embodiment, the following effects are achieved. (1) Since the control rod operations, which were monitored and operated by operators during plant start-up, can be automated, it is possible to save labor in plunge operation and prevent erroneous operations. (2) In addition to changing the amount of control rod operation according to the reactor output and the thermal margin of the reactor core, it also has a function that learns based on changes in the plant state after control rod operation, so the reactor output Unnecessary fluctuations are minimized. This avoids unnecessary plant trips. In the end, it will be possible to start up the reactor output in accordance with the target in a short period of time. (3) Since this is an automation method that utilizes the operator's qualitative knowledge, there is no need for complex control algorithms for automation. In particular, by using fuzzy inference to determine and learn the amount of control rod operation, it is possible to easily achieve functions equivalent to or better than that of an operator who handles many variables. Software development and maintenance will also become easier. In the embodiment of the present invention described above, the description focused on the operation to increase the reactor output at the time of plant start-up. During reactor critical operation or reactor temperature and pressure increase operation, by inputting the reactor period or reactor temperature (reactor water temperature), operation can be realized with these parameters controlled. Therefore, by inputting the reactor period, reactor temperature, etc., control rod operations at plant start-up can be fully automated. What is the effect of this? In particular, it is possible to automatically start up a plant in a short time after load shedding occurs due to a lightning strike or the like. Furthermore, the present invention can also be applied when replacing control rod patterns by manually inputting necessary plant data. In this case as well, there are effects such as labor saving in operation, prevention of erroneous operation, stabilization of i-core furnace output, and increase in load follow-up speed. In this example, the reactor output was used as the main steam flow rate, but even if the reactor output is replaced with the generator output, neutron flux, etc.
The present invention is effective. As another embodiment of the present invention, an automatic control rod operation device can be considered that is applied during a large load follow-up operation using both control rod operation and recirculation flow rate control. this is. In the configuration of FIG. 1, as an input signal, the recirculation flow rate. The configuration follows the target trajectory of re-WI recirculation amount and reactor output. According to this embodiment, there is an effect of realizing automatic operation that can follow the set load at high speed while stably controlling the reactor output, recirculation flow rate, etc. [Effects of the Invention] According to the present invention, the following effects can be obtained. (1) Since control rod operations can be automated, it is possible to save labor and prevent operational errors in nuclear power plants. (2) Since the amount of control rod operation is flexibly changed according to plant conditions, fluctuations in reactor output can be minimized and unnecessary plant trips can be avoided. It also becomes possible to start up the nuclear reactor output in accordance with the target in a short period of time. (3) Control using rules can simplify the development and maintenance of control algorithms and software for automation. In particular, by using Feige inference to determine and learn control rod operations, it is possible to easily achieve functionality equivalent to or better than that of an operator who handles many variables.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the control rod automatic operation device according to the present invention, FIG. 2 is a block diagram showing details of the control rod integrated control device in FIG. 1, and FIG.
FIG. 4 is a flowchart showing the processing of the registration unit; FIG. 4 is a block diagram showing the configuration and operation of the knowledge base and interpreter; FIG.
The figure shows a simplified rule for explaining fuzzy inference, Figures 6, 7, and 8 are flowcharts showing calculation procedures in fuzzy inference, and Figures 9, 10,
Figure 11 is a diagram showing the simulation test results without the learning function, Figures 12 and 13 are Figures 10 and 1.
FIG. 1 is a diagram showing simulation test results showing the effects of the present invention when there is a learning function corresponding to FIG. DESCRIPTION OF SYMBOLS 1... Boiling water reactor, 2... Control rod, 3... Reactor core, 4... Recirculation pump, 5... Main steam pipe,
6... Water supply pipe, 10... Control rod control device, 11...
- Control rod drive device, 12... Control rod drive control device, 2
0...Control rod integrated control device, 21...Control rod operation timing determining section, 22...Control rod operation amount determining section, 23...
#Osho operation amount learning section, 21.1...Data input section, 2
12... Input data file, 213... Occurrence event monitoring/registration unit, 214... Past data and corresponding event file, 215... Knowledge base, 216... Interpreter (inference mechanism), 217...・Operation instruction output section, 2
21... Operation amount determination data file, 222... Operation amount calculation unit, 231... Learning rule and membership function file. 232...Calculation section.

Claims (1)

[Claims] 1. Coordinates of at least one control rod to be withdrawn or inserted among a plurality of control rods inserted into the core of a nuclear reactor, the amount of operation of the control rod, and the timing of operation of the control rod. and a control rod control device that operates the control rod according to the coordinates, operation amount, and operation timing of the control rod output from the control rod overall control device. In the automatic control rod operation device, the control rod integrated control device determines the coordinates of the control rod to be operated based on the control rod operation sequence, and also determines the deviation between at least one of the reactor process quantities and its target value. a control rod operation timing determination unit that determines the operation timing of the control rod based on the control rod; and a control rod operation timing determination unit that determines the operation amount of the control rod based on at least the coordinates of the control rod and the extraction positions of several control rods including the control rod. 1. A control rod automatic operation device comprising: a control rod operation amount determination unit. 2. The control rod automatic operation device according to claim 1, wherein the control rod operation amount determination unit determines the operation amount of the control rod using fuzzy logic. 3. The automatic control rod operating device according to claim 1 or 2, wherein the reactor process amount is at least one of reactor power, reactor temperature, and reactor period. 4. The control rod automatic operation device according to any one of the above claims, characterized in that the control rod operation timing determination unit determines the operation timing of the control rod using a production system. 5. Automatically determine the coordinates of at least one control rod to be withdrawn or inserted among the plurality of control rods inserted into the reactor core, the amount of operation of that control rod, and the timing of operation of that control rod. and a control rod control device that operates the control rod according to the coordinates, operation amount, and operation timing of the control rod output from the control rod overall control device. In the step, the control rod integrated control device determines the coordinates of the control rod to be operated based on the control rod operation sequence, and determines the coordinates of the control rod to be operated based on the deviation between at least one of the reactor process quantities and its target value. and a control rod operation amount that determines the operation amount of the control rod based on at least the coordinates of the control rod and the extraction positions of several control rods including the control rod. a determination unit and a control rod operation amount learning unit that corrects part of the data used to determine the control rod operation amount based on the change in the state of the reactor after operating the control rod or the time interval between control rod operations. An automatic control rod operating device comprising: 6. The control rod automatic operation device according to claim 5, characterized in that the control rod operation amount learning section modifies a part of the data used to determine the control rod operation amount using fuzzy logic. . 7. In claim 5 or 6, the state change of the reactor is defined as a deviation between the reactor output and its target value, a deviation between the reactor temperature and its target value, a reactor period and its target value A control rod automatic operation device characterized in that the change is at least one of a deviation from the neutron flux, a neutron flux, and a thermal margin of a reactor core. 8. A control rod according to any one of claims 5 to 7, characterized in that the control rod operation timing determination unit determines the operation timing of the control rod using a production system. Automatic operating device.