CN106226055B - A kind of monitoring reliability method of nuclear power plant's valve body failure based on fault tree - Google Patents

Abstract

The invention belongs to the field of nuclear power plant risk monitor (Risk Monitor), and in particular relates to a failure tree-based reliability of valve body failure of a nuclear power plant based on fault tree modeling technology, which is suitable for failure of valve equipment body failure in online risk monitoring of nuclear power plants monitoring method. The invention includes: (1) online collection of state monitoring information of nuclear power plant valves; (2) online recognition of the initial state of the valve and establishment of an initial state transfer diagram; (3) establishment of a modular fault tree reliability model for failure of the valve equipment body and Online update; (4) Online calculation of the failure reliability of the valve equipment body based on the fault tree. Based on the fault tree method and state monitoring technology, the present invention provides a framework and steps of a reliability monitoring method for the failure of the valve equipment body in a nuclear power plant, which can overcome the limitations of the traditional fault tree method in terms of modeling valve equipment status and online update insufficient.

Description

A Reliability Monitoring Method of Valve Body Failure in Nuclear Power Plant Based on Fault Tree

technical field

The invention belongs to the field of nuclear power plant risk monitor (Risk Monitor), and in particular relates to a failure tree-based reliability of valve body failure of a nuclear power plant based on fault tree modeling technology, which is suitable for failure of valve equipment body failure in online risk monitoring of nuclear power plants monitoring method.

Background technique

Nuclear power plant risk monitoring technology is the specific application and development of probabilistic safety assessment (PSA) technology in the process of nuclear power plant operation. The use of risk monitoring technology in nuclear power plants can realize the known and controllable risks of nuclear power plants, so as to improve the economical efficiency of nuclear power plant operation under the premise of ensuring safety. When carrying out risk monitoring of nuclear power plants, it is necessary to consider the power plant configuration, that is, the combination of the availability status of various equipment and systems. A large number of valve devices are used in nuclear power plants. During the operation of the nuclear power plant, the valve may be in different states, such as open or closed, and the state of the valve also changes with time, which will make its risk analysis complicated.

At present, in the risk monitoring of nuclear power plants, the fault tree analysis method is generally used to analyze the reliability of the system and equipment. In traditional nuclear power plant risk monitoring, the focus is on evaluating the risks of expected planning activities such as maintenance and testing. Usually, for some special power plant configurations, fault tree configuration exhaustive modeling is carried out through room-shaped events to meet specific group requirements. System and equipment reliability modeling requirements under the state. For valve equipment, it is usually to specify its initial state under a certain configuration and combine its demand state to analyze the relevant failure modes and establish a fault tree model. Taking the electric valve RRI040VN/041VN in the cooling water system of a nuclear power plant as an example, the analysis of traditional risk monitoring is shown in Table 1 below. The schematic diagram of VN electric isolation valve in nuclear power plant is shown in Figure 1.

Table 1 Information sheet of electric valve RRI040/041 for cooling water system in nuclear power plant

For configuration 1 in Table 1, the established fault tree reliability models of RRI040VN and RRI041VN are shown in Figure 2 and Figure 3.

Similarly, the corresponding fault tree reliability model can also be established for the configuration 2 in table a, and the corresponding configuration can be selected through the room shape event.

With the gradual application of computer technology and digital instrument control system in nuclear power plants, the risk monitoring technology of nuclear power plants has also been further developed, and the concept of online risk monitoring has been proposed. The so-called online risk monitoring is to collect the operating data of nuclear power plants through state monitoring technology, such as system operating pressure, flow, valve switch and other measurable information, automatically identify the status of nuclear power plant systems and equipment, and use the online risk model to timely Evaluate the current risks of nuclear power plants. The online risk model requires the configuration of the power plant to be flexibly modeled, and has the characteristics of timely update and fast calculation. However, the traditional reliability modeling method of valve equipment body failure based on fault tree and room shape events has the following deficiencies:

(1) Insufficient flexibility. In the actual operation of nuclear power plants, there are often switching states of valves or other dynamic processes related to valve actions. The traditional method of controlling configuration through room-shaped events usually only models the power plant configuration before and after valve action under typical working conditions, and the influence of the valve action process is generally not modeled. This does not meet the needs of online risk monitoring.

(2) It is not convenient to update the valve status. In the traditional analysis method, the specific configuration of the valve equipment and the corresponding valve initial state/demand state are generally stipulated, and the typical valve state is analyzed through state exhaustion. This cannot effectively use the collected information to achieve a more convenient and continuous update of the model under the background of the existing nuclear power plant informatization.

Therefore, in the on-line risk monitoring of nuclear power plants, it is necessary to establish a flexible and effective reliability monitoring method for valve equipment to replace the traditional valve equipment reliability model based on fault tree and room-shaped events. There are many different methods for the study of valve equipment reliability modeling methods. However, in the field of practical nuclear power engineering, the fault tree analysis method is still the main method, and the mainstream calculation software is also based on the fault tree calculation. Based on this, the present invention proposes a set of valve equipment body failure reliability monitoring method based on fault tree on the basis of nuclear power plant-related valve state monitoring technology, which strengthens the update capability of the valve equipment body failure reliability model, and provides a comprehensive solution for nuclear power plants. The modeling of valve equipment in online risk monitoring provides a new method.

Contents of the invention

The purpose of the present invention is to provide a reliability monitoring method for failure of a nuclear power plant valve body based on a fault tree.

The purpose of the present invention is achieved like this:

A method for reliability monitoring of failure of a nuclear power plant valve body based on a fault tree, comprising the following steps:

(1) Online collection of status monitoring information of nuclear power plant valves:

Collect monitoring signals of valve equipment from the information monitoring and management system of the nuclear power plant, specifically including collecting valve switch signals from the real-time monitoring system, collecting valve test/maintenance isolation record information from the auxiliary isolation computing system, and collecting valve data from the digital control system. Start the control signal; according to the above signals and information, construct a condition monitoring symptom space Ω of the valve equipment. The elements contained in Ω are: valve normally open signal (1), valve normally closed signal (0), valve maintenance/fault record (r ), the valve open to close step control signal (-1), the valve close to open step control signal (+1); among them, the valve normally open signal and normally closed signal are collectively referred to as the valve constant signal; the valve open to close and close to open The step control signal is collectively referred to as the valve step signal; the valve equipment status monitoring symptom space Ω is expressed as:

Ω=(1,0,r,-1,+1);

(2) Identify the initial state of the valve online and establish an initial state transition diagram;

(2.1): Establish the initial state space S _o of valve equipment; S _o includes states: open (s ₁ ), closed (s ₂ ), fault (s ₃ ), preventive maintenance (s ₄ ), corrective maintenance (s ₅ ); the initial state space S _o of the valve equipment is expressed as:

S _O ＝(s ₁ ,s ₂ ,s ₃ ,s ₄ ,s ₅ )

Through the state monitoring symptom space Ω of the valve equipment, it is possible to monitor which initial state the valve is in in the state space S _o during actual operation, and determine the criterion function f _C from the valve monitoring symptom space Ω to the initial state space S _o of the valve equipment, Methods as below:

(2.1.1) When the valve constant signal is 1, the state is open;

(2.1.2) When the valve constant signal is 0, the state is off;

(2.1.3) Record information to determine the status as corrective maintenance, preventive maintenance or failure;

(2.1.4) The valve step signal +1 or -1 represents the automatic start of the valve state open or close;

Establish the following mapping relationship:

(2.2) Establish the initial state transition diagram of the valve according to the actual operation; the establishment method of the initial state transition diagram is as follows:

(2.2.1) Determine the initial state of the valve based on the established initial state space S _O of the valve equipment, represented by an oval box;

(2.2.2) Determine the transfer path between the initial states based on the actual operation of the power plant, represented by a solid line with arrows and numbers, and the number indicates the sequence number of the transfer path between the initial states;

The determined complete transfer path includes: during normal operation, the mutual conversion between opening (s ₁ ) and closing (s ₂ ); during maintenance, the mutual conversion between opening (s ₁ ) and preventive maintenance (s ₃ ) and closing (s ₂ ) and preventive maintenance (s ₃ ); in the process of failure, the transition from open (s ₁ ) to fault (s ₃ ) and from off (s ₂ ) to fault (s ₃ ); in the process of repair, the fault (s ₃ ) to corrective maintenance (s ₅ ), corrective maintenance (s ₅ ) to open (s ₁ ) and corrective maintenance (s ₅ ) to close (s ₁ ) transitions; valve equipment condition monitoring symptom space Ω and the criterion function f _C monitor the initial state or change, and add relevant monitoring signal input in the state transition diagram;

(3) Establish a modular fault tree reliability model of valve equipment body failure and update it online:

(3.1) Establish the initial state-demand state transition diagram of the valve equipment:

Establish the demand state space S _d of the valve equipment, and the states included in S _d are: open (S _d1 ), closed (S _d2 ), and no demand (O);

S _d = (s _d1 ,s _d2 ,o);

Establish the mapping relationship from the initial state of the valve equipment to the demand state, and establish the rule f _s as follows:

(3.1.1) "initial state" selects the elements in the defined initial state space S ₀ ;

(3.1.2) "Demand state" selects elements in the defined demand state space S _d ;

(3.1.3) When the "initial state" of the valve equipment is determined to be unavailable, the "demand state" is empty;

(3.1.4) When the "initial state" of the valve equipment is determined to be available, determine the mapping object according to the actual operation;

Based on the rule f _s , the mapping relationship from the initial state of the valve equipment to the demand state is: when the valve equipment is initially in the open (S ₁ ) state, the power plant is specified as available, and all demands include the open (S _d1 ) and closed (S _d2 ) states; When the valve is initially in the closed (S ₁ ) state, the power plant is specified as available, and all requirements include the open (S _d1 ) and closed (S _d2 ) states; the valve equipment is initially faulty (S ₃ ), preventive maintenance (S ₄ ), corrective In the maintenance (S ₅ ) state, the power plant is defined as unavailable, and the demand state is empty at this time, which corresponds to "no demand (O). The established mapping relationship is as follows:

Based on the initial state transition diagram, a complete initial state-demand state transition diagram of valve equipment is established, and the establishment rules are:

(3.1.1.1) Determine the demand state of the valve based on the established valve equipment demand state space _Sd , which is represented by a box;

(3.1.1.2) Determine the transition path from the initial state to the demand state based on the actual operation of the power plant, represented by a solid line with arrows and letters, and the letter represents the sequence number of the transition path from the initial state to the demand state;

Through the mapping relationship between the initial state of the valve equipment and the demand state, according to the establishment rules, a complete initial state-demand state transition diagram of the valve equipment is established on the monitoring and transition diagram of the initial state of the valve in the nuclear power plant;

(3.2) Establish a modular fault tree reliability model for failure of the valve equipment body:

The modeling method of valve equipment modular fault tree is as follows:

(3.2.1) The modular fault tree does not define the initial state of the equipment;

(3.2.2) Modular fault tree envelops the failure modes of all analyzed states in the equipment state transition diagram;

After the modular fault tree model of the equipment is established, the minimum cut set set of the fault tree is solved by Boolean operation; if it contains N minimum cut sets, the jth minimum cut set is expressed as:

K _j ＝(X ^j ₁ ,X ^j ₂ ,…,X ^j _i ,····,X ^j _n )

X ^j _i represents the failure mode, and the minimum cut set set of modular fault tree is expressed as:

Θ={K ₁ ,K ₂ ,...,K _N }

In Θ, as long as all the bottom events X ^j _i of any minimum cut set K _j (j=1,2,…,N) occur, the top event of the fault tree must occur, and the minimum cut set set is used to represent the fault tree Structural function; each fault tree structural function is the result of simplification and absorption after the set operation of the fault tree, that is, the modular fault tree structural function Θ ₀ (X) is expressed as:

The failure modes included in various valves are as follows:

(3.2.1.1) Electric valve: rupture, external leakage, internal leakage, refusal to open, refusal to close, wrong opening, wrong closing;

(3.2.1.2) Check valve: rupture, external leakage, internal leakage, refusal to open;

(3.2.1.3) Manual valve: rupture, external leakage, internal leakage, forget to open, forget to close;

(3.2.1.4) Pneumatic valve: rupture, external leakage, internal leakage, wrong closing, wrong closing, manual opening failure;

When the valve equipment is in the initial state of "preventive maintenance", "corrective maintenance" and "fault", it means that it is unavailable. At this time, only fault/maintenance unavailable failure modes are included, and the unavailable logic update processing of the fault tree is performed;

(3.3) Establish the update rules of the valve equipment modular fault tree, and obtain the mapping relationship between the modular fault tree and each state fault tree:

(3.3.1) If the X ^j _i event attribute in the modular fault tree is set to False, it means that the event will not happen; then remove all the minimum cut sets containing X ^j _i in the fault tree structure function Θ ₀ (X) , re-perform the set operation to absorb and simplify to generate a new structure function Θ _i (X);

(3.3.2) If the X ^j _i event attribute in the modular fault tree is set to True, it means that the event has occurred; remove the X ^j _i elements contained in the minimum cut set in the fault tree structure function Θ ₀ (X), and re- After performing set operation absorption and simplification, a new structural function Θ _i (X) is generated;

(3.3.3) If the X ^j _i event attribute in the modular fault tree is set to Normal, which means that the event occurs with a certain probability, then it is substituted into the basic event reliability model for calculation; that is, the fault tree structure function Θ ₀ (X) does not Change;

On the basis of the valve modular fault tree model, the fault tree model of each state is obtained; that is, through the modular fault tree structure function Θ ₀ (X), the actual mapping relationship f _x is established, and the structural function of the fault tree model of each state of the equipment is obtained ; The mapping is expressed as:

For each determined valve equipment state, there is a corresponding state fault tree structure function Θ _i (X):

Θ _i (X) = f _x ⁱ [Θ ₀ (X)]

The secondary update of the characteristics for the valve demand status change is as follows:

(3.3.3.1) One-time update: when the initial state of the valve is monitored, the demand state is set to be the same as the initial state, and the valve state is updated; the initial state of the valve is monitored to be unchanged, and the power plant configuration changes cause the valve demand state to be updated, and the valve state is directly updated. status updates;

(3.3.3.2) Secondary update: For the change of the initial state of the valve detected in one update, judge whether the actual demand state of the valve is the same as the initial state. Second update; establish the second update rule of modular fault tree:

When the valve state is updated each time, define a transfer process from the F _i state to the F _j state:

F _i→j ＝F _j -F _i

F _i→j corresponds to the transition path in the state transition diagram; in Θ(X), there is a fault tree structure function Θ _i (X) of F _i state to the fault tree structure function Θ _j (X) of F _j state The change process:

Θ _i→j (X)＝Θ _j (X)-Θ _i (X)

The generated Φ _i→j (X) is the update rule, and has the following correspondence:

F _i→j ——→Φ _i→j (X)

In the valve body failure modular fault tree bottom event, rupture and external leakage can be summarized as open/close operation failure; internal leakage is demand-close operation failure; refusal to open, refusal to close, mis-open, and mis-close can be summarized as demand failure; Among them, refusal to open and false opening are the process of closing to opening, which is called demand failure 1; refusal to close and wrong closing are the process of opening to closing, called demand failure 2;

One update path: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, a, d path;

Secondary update path: 1-d, 2-a, 6-a, 7-d, 9-a, 11-d path;

(3.4) Judging the state of the valve and executing update rules for its modular fault tree; first judge the "initial state" of the valve, if it does not change, judge the demand state of the valve, and if it changes, perform the first update of its modular fault tree and then judge Valve demand status; then judge the valve demand status under the condition that the initial status remains unchanged. If it remains unchanged, it is not necessary to update its modular fault tree. state, if it remains unchanged, there is no need to perform a second update on its modular fault tree, and if it changes, a second update is required;

(4) On-line calculation of the reliability of the failure of the valve equipment body based on the fault tree;

Establish a modular fault tree for the failure of the valve body, and assign logical values to the modular fault tree by calling the established update rules; simplify and absorb the Boolean set operation of the fault tree model to obtain the minimum cut set set of the reliability model; finally transform For the corresponding fault tree structure function, after disjoint processing, it is substituted into the corresponding basic event reliability model, and the failure probability of valve body failure is calculated; the structure function Φ(X) transformed by the minimum cut set set is:

Substituting the failure probability P ⁶ _i (t) of the basic event obtained by the reliability model of the basic event X ^j _i at time t, the unreliability R(X,t) of the failure of the valve equipment body is obtained, namely:

The beneficial effects of the present invention are:

(1) The present invention is based on the fault tree method and state monitoring technology, provides a kind of framework and the step of the reliability monitoring method for the valve equipment body failure of nuclear power plant, can overcome traditional fault tree method in modeling valve equipment state and on-line Lack of updates.

(2) The present invention provides a basic modeling idea for the failure of the valve equipment body in the online risk model of the nuclear power plant, considers multiple state situations, effectively reduces the scale of the valve logic fault tree and simplifies the modeling process, and reduces the number of models development costs.

(3) The present invention has good compatibility with existing widely used commercial software (such as RiskSpectrum), can make full use of existing fault tree-based modeling software, is easy to be accepted by nuclear power engineering and application personnel, and is convenient for engineering realization.

Description of drawings

Figure 1 Initial state monitoring and state transition diagram of valve equipment in a nuclear power plant;

Figure 2 The fault tree reliability model established by configuration 1 for RRI040VN;

Figure 3 The fault tree reliability model established by configuration 2 for RRI041;

Figure 4 Initial state monitoring and state transition diagram of valve equipment in nuclear power plant;

Fig. 5 "initial state-demand state" transition diagram of nuclear power plant valve equipment;

Fig. 6 The reliability monitoring process of valve equipment body failure based on fault tree in online risk monitoring of nuclear power plant;

Fig. 7 Logic model of modular fault tree for failure of RRI040VN electric valve in nuclear power plant.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings.

The invention provides a reliability monitoring method for the failure of the valve equipment body of a nuclear power plant, which is used to establish a failure reliability model of the valve equipment body in the online risk monitoring of the nuclear power plant, improves the dynamic update capability of the reliability model of the valve equipment body failure, and overcomes the The deficiencies of the traditional risk monitoring model in the flexible modeling of valve state changes have been realized, and the continuous monitoring of the failure of the valve equipment in nuclear power plants has been realized.

In order to achieve the above object, the technical solution of the present invention is achieved in that:

Step 1: Online collection of status monitoring information for nuclear power plant valves.

The invention collects the monitoring signal of the valve equipment from the information monitoring and management system of the nuclear power plant, specifically including collecting the switch signal of the valve from the real-time monitoring system (KNS), and collecting the valve test/maintenance isolation record information from the auxiliary isolation calculation system (CBA) , collect the start control signal of the valve from the digital control system (DCS). According to these signals and information, a condition monitoring symptom space Ω of valve equipment can be constructed. The elements contained in Ω are: valve normally open signal (1), valve normally closed signal (0), valve maintenance/fault record (r), valve Open to close step control signal (-1), valve close to open step control signal (+1). Among them, the valve normally open signal and normally closed signal are collectively referred to as the valve constant signal; the valve open-to-close and close-to-open step control signals are collectively referred to as the valve step signal. Therefore, the valve equipment status monitoring symptom space Ω can be expressed as:

Ω＝(1,0,r,-1,+1)

The present invention constructs a valve state monitoring symptom space Ω, with the purpose of screening out monitoring signals and information actually related to the update of the failure reliability model of the valve body from the massive monitoring information of the nuclear power plant.

Step 2: Identify the initial state of the valve online and establish an initial state transition graph.

Step 1: The present invention establishes the initial state space S _o of the valve equipment based on the requirements of the actual operation of the nuclear power plant and the failure reliability analysis of the valve body in the online risk model. In the present invention, the states included in S _o are: open (s ₁ ), closed (s ₂ ), fault (s ₃ ), preventive maintenance (s ₄ ), and corrective maintenance (s ₅ ). The initial state space S _o of the valve equipment can be expressed as:

S _O ＝(s ₁ ,s ₂ ,s ₃ ,s ₄ ,s ₅ )

Through the establishment of the state monitoring symptom space Ω of the valve equipment, the present invention requires monitoring which initial state the valve is in in the state space S _o in actual operation, and it is necessary to determine the relationship between the valve monitoring symptom space Ω and the initial state space S _o of the valve equipment Criterion function f _C . The regulations are as follows:

(1) When the valve constant signal is "1", the state is "open";

(2) When the valve constant signal is "0", the state is "off";

(3) Record information and determine the status as "corrective maintenance", "preventive maintenance" or "failure";

(4) The valve step signal "+1" or "-1" represents the automatic start of the valve state "open" or "close".

Therefore, the following mapping relationship is established:

Step 2: The present invention needs to establish a valve initial state transition diagram according to actual operation. The initial state transition diagram provides the basis for the initial state update of the failure reliability model of the valve equipment body. The rules for establishing the initial state transition diagram specified in the present invention are as follows:

(1) Determine the initial state of the valve based on the established initial state space S _O of the valve equipment, represented by an oval box;

(2) Determine the transfer path between the initial states based on the actual operation of the power plant, which is represented by a solid line with arrows and numbers, and the number indicates the sequence number of the transfer path between the initial states.

The present invention defines five initial states of the valve equipment in the state space S _O , including: open (s ₁ ), close (s ₂ ), failure (s ₃ ), preventive maintenance (s ₄ ), corrective maintenance (s ₅ ). Based on the actual operation process of the nuclear power plant, the complete transfer path determined includes: during normal operation, the mutual conversion between on (s ₁ ) and off (s ₂ ); during maintenance, on (s ₁ ) and preventive maintenance (s ₃ ) Mutual conversion of switching off (s ₂ ) and preventive maintenance (s ₃ ); during failure process, switching from switching on (s ₁ ) to fault (s ₃ ) and switching from closing (s ₂ ) to fault (s ₃ ) Transitions; during repair, transition from failure (s ₃ ) to corrective maintenance (s ₅ ), transition from corrective maintenance (s ₅ ) to on (s ₁ ), and transition from corrective maintenance (s ₅ ) to off (s ₁ ) . In addition, the present invention can monitor its initial state or change through the above-mentioned valve equipment state monitoring symptom space Ω and criterion function f _C , and can also add relevant monitoring signal input in the state transition diagram (the circle with shadow can be used for monitoring symptoms) shape, and the input path is indicated by a dotted line with an arrow).

Therefore, the monitoring and transition diagram of the initial state of the nuclear power plant valve established by the present invention is shown in Figure 4, from which it can be seen that there are at most 11 state transition paths in the initial state of the nuclear power plant valve.

The initial state transition diagram provides the basis for the initial state update of the valve equipment reliability model.

Step 3: Establish a modular fault tree reliability model of valve equipment body failure and update it online.

Step 1: Establish the "initial state - demand state" transition diagram of the valve equipment. In the fault tree analysis method, the state of the valve equipment is determined by the "initial state" and "demand state". In the present invention, the initial state space S ₀ of the valve device and the monitoring and transition diagram of the initial state (Fig. 4) have been established, and the "initial state" of the valve device can be monitored.

However, the "demand state" in the fault tree analysis of valve equipment body failure has only two states: "open" and "closed", which are related to the actual configuration of the power plant and are difficult to monitor automatically. "Demand state" exists only when the "initial state" of the valve is normal "operating" or "standby" (may be "closed" or "open" state), if the valve is not in service due to "maintenance" or "failure" Typically there is no "requirement state". Therefore, the present invention also establishes the demand state space S _d of the valve equipment, and the states included in S _d are: open (S _d1 ), closed (S _d2 ), and no demand (O).

S _d =(s _d1 ,s _d2 ,o)

In order to clarify the change model of the valve equipment state, combined with the actual operation of the valve equipment in the nuclear power plant, the present invention establishes the mapping relationship from the initial state of the valve equipment to the demand state, and the establishment rule f _s is as follows:

(1) "initial state" selects the elements in the defined initial state space S ₀ ;

(2) "Demand state" selects elements in the defined demand state space S _d ;

(3) When the "initial state" of the valve equipment is determined to be unavailable, the "demand state" is empty.

(4) When the "initial state" of the valve equipment is determined to be available, the mapping object is determined according to the actual operation.

Based on the rule f _s , the mapping relationship from the initial state of the valve equipment to the demand state is: when the valve equipment is initially in the "open (S ₁ )" state, the power plant is specified as available, and all demands include "open (S _d1 )" and "closed ( S _d2 )" state (such as electric isolation valve); when the valve is initially in the "closed (S ₁ )" state, the power plant is specified as available, and all requirements include "open (S _d1 )" and "closed (S _d2 )" states ( Such as manual isolation valve); when the valve equipment is initially in the state of "fault (S ₃ )", "preventive maintenance (S ₄ )", "corrective maintenance (S ₅ )", the power plant is specified as unavailable, and the demand state is empty at this time , corresponding to "no demand (O)". Therefore, the established mapping relationship can be expressed as follows:

After the above mapping is determined, the present invention establishes a complete "initial state-demand state" transition diagram of valve equipment on the basis of the initial state transition diagram, and adds the following establishment rules:

(1) Determine the demand state of the valve based on the established valve equipment demand state space _Sd , which is represented by a box;

(2) Determine the transition path from the initial state to the demand state based on the actual operation of the power plant, which is represented by a solid line with arrows and letters, and the letter represents the sequence number of the transition path from the initial state to the demand state.

Through the above-mentioned mapping relationship from the initial state of the valve equipment to the demand state, according to the established rules added, the present invention can establish a complete "initial state-demand state" transition diagram of the valve equipment on the monitoring and transition diagram of the initial state of the valve equipment in the nuclear power plant , as shown in Figure 5.

Step 2: Establish a modular fault tree reliability model for valve equipment body failure. The fault tree model is a graphical logic model, and the failure reliability of the valve equipment body can be obtained by combining the fault tree logic model with the basic event reliability model. The invention proposes a modularized fault tree method to solve the problem of difficulty in updating fault tree models in the case of multi-state combination of valve equipment. The idea of modular fault tree modeling for valve equipment is as follows:

(1) The modular fault tree does not define the initial state of the equipment;

(2) The modular fault tree envelops the failure modes of all analyzed states in the equipment state transition diagram.

After the modularized fault tree model of the equipment is established, the minimum cut set set of the fault tree can be obtained through Boolean operations. If there are N minimum cut sets, the jth minimum cut set is expressed as:

K _j ＝(X ^j ₁ ,X ^j ₂ ,…,X ^j _i ,····,X ^j _n )

X ^j _i represents the failure mode (primary event or bottom event). Then, the minimum cut set set of modular fault tree is expressed as:

Θ={K ₁ ,K ₂ ,...,K _N }

In θ, as long as all the bottom events X ^j _i of any minimum cut set K _j (j=1,2,…,N) occur, the top event of the fault tree must occur. Therefore, the minimum cut set can be used to represent the structural function of the fault tree. And each fault tree structure function is the result of simplification and absorption after the set operation of the fault tree. That is, the modular fault tree structure function Θ ₀ (X) can be expressed as (mathematical form of modular fault tree):

The present invention is based on the idea of modular fault tree modeling of the above-mentioned valve equipment, and establishes the failure reliability model of the valve equipment body based on the modular fault tree as follows:

The modular fault tree of valve equipment does not specify the valve state, but it needs to envelop the failure modes of the body failure in all states of the valve in the nuclear power plant. Generally, nuclear power plant valves include electric valves, check valves, manual valves and pneumatic valves. When only the failure of the valve body is considered, the failure modes included in various valves are as follows:

(1) Electric valve: rupture, external leakage, internal leakage, refusal to open, refusal to close, wrong opening, wrong closing;

(2) Check valve: rupture, external leakage, internal leakage, refusal to open;

(3) Manual valve: rupture, external leakage, internal leakage, forget to open, forget to close;

(4) Pneumatic valve: rupture, external leakage, internal leakage, wrong closing, wrong closing, manual opening failure.

In the present invention, when the valve equipment is in the initial state of "preventive maintenance", "corrective maintenance" and "fault", it means that it is unavailable. At this time, only the failure mode of "fault/maintenance unavailable" is included, and only the unavailability of the fault tree is needed. Handle with logical updates.

Therefore, for the above four types of nuclear power plant valves, the present invention can respectively establish corresponding modular fault trees (connected by OR gates) according to the idea of fault tree analysis.

Step 3: The present invention establishes the update rule of the valve equipment modular fault tree. After the modular fault tree is established, it is necessary to establish the mapping relationship from the modular fault tree model to each state fault tree model, that is, update rules. The idea of updating the modular fault tree of the equipment is as follows:

Since the equipment-level modular fault tree model is a bottom-layer graphical logic model, and the bottom layer covers the failure modes (bottom events) of all analyzed states of the equipment, the modular fault tree also covers each state fault tree. Due to the convenience of the graphical model, the mapping relationship between the modular fault tree and each state fault tree can be directly obtained through the following mapping rules:

(1) If the X ^j _i event attribute in the modular fault tree is set to "False", it means that the event will definitely not happen; then remove all the minimum cut sets containing X ^j _i in the fault tree structure function Θ ₀ (X) , re-perform the set operation to absorb and simplify to generate a new structure function Θ _i (X).

(2) If the X ^j _i event attribute in the modular fault tree is set to "True", it means that the event has occurred; then remove the X ^j _i elements contained in the minimum cut set in the fault tree structure function Θ ₀ (X), Re-perform the set operation to absorb and simplify to generate a new structure function Θ _i (X).

(3) If the X ^j _i event attribute in the modular fault tree is set to "Normal", it means that the event occurs with a certain probability, and it can be substituted into the basic event reliability model for calculation; that is, the fault tree structure function Θ ₀ (X )not changing.

Through the above mapping rules, on the basis of the valve modular fault tree model (without initial state), the fault tree model of each state can be obtained. That is, through the modularized fault tree structure function Θ ₀ (X), the actual mapping relationship f _x is established (according to the mapping rules), and the structural functions of the fault tree models of each state of the equipment can be obtained. The mapping is expressed as:

For each determined valve equipment state, there is a corresponding state fault tree structure function Θ _i (X), and there is:

Θ _i (X) = f _x ⁱ [Θ ₀ (X)]

In fault tree analysis, the state of valve equipment is determined by "initial state" and "demand state". Not only the "initial state" will cause the valve state to be updated (as shown in Figure 4), but the change of the "demand state" will also cause the valve state to be updated. Unlike other equipment with a single demand state (such as the demand state of a pump is only running), the "demand state" of the same valve will be different in different functional fault trees. For example, in a nuclear power plant, the common column B column isolation valve RRI040VN of the cooling water system (initial state is "closed") in the fault tree of "RRIA column fails to supply water to the public circuit" demand state is "closed", and in the "RRIB column supplies water to the public circuit" The requirement state in the failure tree is "open". Therefore, although the "initial state" of the same valve is the same, because the "requirement state" will be different in different functional fault trees, the established state fault tree model will also change.

The present invention proposes a secondary update scheme aiming at the characteristics of change and update of the "demand state" of the valve. The scheme is as follows:

(1) One update: ① The change of the "initial state" of the valve is detected. At this time, it is assumed that the "demand state" is the same as the "initial state", and the valve state is updated. ② It is detected that the "initial state" of the valve remains unchanged, and the change of the power plant configuration causes the "demand state" of the valve to be updated. At this point, the valve status is updated directly. The above needs to build a modular fault tree with one update rule.

(2) Second update: For the situation ① in an update, it is judged at this time whether the actual "demand state" of the valve is the same as the "initial state". Status" is updated twice. The above needs to establish a secondary update rule for the modular fault tree.

Every time the valve state ("initial state" or "demand state") is updated, a transition process from the F _i state to the F _j state is defined:

F _i→j ＝F _j -F _i

F _i→j corresponds to the "transition path" in the "state transition diagram". Correspondingly, in Θ(X), there is a fault tree structure function Θ _i (X) of F _i state to the change process of the fault tree structure function Θ _j (X) of F _j state:

Θ _i→j (X)＝Θ _j (X)-Θ _i (X)

Then, the generated Φ _i→j (X) is the update rule established by the present invention. and have the following correspondences:

F _i→j ——→Φ _i→j (X)

In the valve body failure modular fault tree bottom event of the present invention, rupture and external leakage can be summarized as on/off operation failure (valve body failure, RF for short); internal leakage is demand-off operation failure (IL for short). Refusing to open, refusing to close, mistakenly opening, and mistakenly closing can be summarized as demand failure; among them, "refusal to open" and "mistaken opening" are the process from closing to opening, which are called demand failure 1 (NF1 for short); "Close" is the process from open to close, which is called demand failure 2 (abbreviated as NF2).

For each state update path in Fig. 5, the present invention needs to establish the first and second update rules according to the previous modular fault tree update idea and the above-mentioned three types of failure modes (RF, NF1 and NF2). Specifically, 17 update rules are involved, including:

(1) One update path: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, a, d path;

(2) Secondary update paths: 1-d, 2-a, 6-a, 7-d, 9-a, 11-d paths;

Step 4: Determine the state of the valve and execute update rules on its modular fault tree. The present invention can automatically identify the "initial state" of the valve through the state monitoring technology, but needs to directly judge the "demand state" of the valve. According to the above-mentioned secondary update scheme: first judge the "initial state" of the valve, if it remains unchanged, judge the "demand state" of the valve, and if it changes, execute the first update of its modular fault tree and then judge the "demand state" of the valve; then When the "initial state" is unchanged, the "demand state" of the valve is judged. If it is unchanged, there is no need to update its modular fault tree. If it changes, the first update is also performed; finally, the valve is judged when the "initial state" changes. "Requirement status", if it is unchanged, it is not necessary to perform a second update on its modular fault tree, and if it changes, a second update is required.

The fourth step: online calculation of the failure reliability of the valve equipment body based on the fault tree.

The present invention can establish a modular fault tree for failure of the valve body, and carry out logical assignment to the modular fault tree by invoking the established update rules; then obtain the minimum cut set of the reliability model by simplifying and absorbing the Boolean set operation of the fault tree model Finally, it is transformed into the corresponding fault tree structure function, which can be substituted into the corresponding basic event reliability model after approximate disjoint processing to calculate the failure probability of valve body failure. The structural function Φ(X) transformed by the minimum cut set set is:

Therefore, by substituting the basic event failure probability P ^j _i (t) obtained by the basic event X ^j _i reliability model at time t, the unreliability R(X,t) of valve equipment body failure can be obtained, namely:

The failure reliability model of the valve equipment body in the present invention adopts the basic reliability model of exponential distribution widely used in the safety analysis of nuclear power plants. The unreliability R(X,t) of the failure of the valve equipment body is obtained, which is finally used as the bottom layer input of the online risk monitoring model of the nuclear power plant.

Figure 6 shows the reliability monitoring process of valve equipment body failure based on fault tree in online risk monitoring. The present invention will be described by taking the electric valve RRI040VN of the cooling water system of the nuclear power plant as an example.

The initial state monitoring and state transition diagram of the valve equipment in the nuclear power plant shown in Figure 1, combined with the daily actual production management of the nuclear power plant, the initial state of the plant with the valve equipment in the nuclear power plant is shown in Table 2 below:

Table 2 Initial status of valve equipment in nuclear power plants

Based on the status monitoring symptoms and criteria of the valve equipment, the "initial status" monitoring analysis of the RRI040VN electric valve in the nuclear power plant is shown in Table 3 below.

Table 3 Symptom monitoring analysis of initial state of RRI040VN electric valve in nuclear power plant

Based on the analysis of Table 2 and Table 3, the "initial state" monitoring and transition diagram of the RRI040VN electric valve established by the method proposed by the present invention refers to Figure 4, and the "initial state-demand state" transition diagram of the established RRI040VN electric valve refers to the diagram 5.

Based on the failure mode analysis of the valve equipment, the failure mode analysis of the nuclear power plant RRI040VN electric valve under all states is shown in Table 4 below.

Table 4 Classification of failure modes of one-time change of RRI040VN electric valve in nuclear power plant

Based on the analysis in Table 4, the failure modes of the valve equipment are summarized, and a modular fault tree logic model for the failure of the RRI040VN electric valve body is established, as shown in Figure 7.

According to the "initial state-demand state" transition diagram of nuclear power plant valve equipment shown in Figure 5, there are 17 change paths according to the actual operation, and the description of the state change process and related change paths is shown in Table 5 below. The change process in the table Inside the parentheses is the requirement status.

Table 5 Actual change process of NPP RRI040VN

In combination with the secondary update scheme proposed by the present invention, the update rule Φ _i→j (X) of the secondary update scheme established is shown in Table 6 below.

Table 6 Secondary update rules for modular fault tree of nuclear power plant RRI040VN electric valve

The failure mode (bottom event) of all states is included in the valve equipment modular fault tree and the initial assignment defaults to "Normal". However, valves contain different failure modes under different conditions.

Therefore, based on the established mapping rules, the present invention assigns values ("False", "True", and "Normal") to the corresponding failure modes in the modular fault tree to accurately model the logical models of the equipment in different states. When the valve is in a certain state, the failure modes not included in the modular fault tree are assigned the value "False", and the failure modes included in the modular fault tree remain "Normal". Moreover, the "unavailable" failure mode exists in the fault tree only in the unavailable state, and it indicates that the event has occurred, and the value "True" is assigned; while in other states, the "unavailable" failure mode does not exist in the fault tree , thus assigning the value "False".

The valve equipment body failure reliability monitoring model proposed above is applicable to all functional systems including valve equipment in nuclear power plants, and can provide a basic reliability model for valve equipment body failure reliability for related functional systems. The actual operation mode of the failure reliability monitoring model of the valve equipment body based on the fault tree in the online risk monitoring is as follows: first, the risk monitor collects the monitorable symptoms of the valve equipment analyzed in Table 3, and identifies each one online through the set symptom criteria. The state of the valve; then, through the defined secondary update scheme, the initial state of the valve equipment is automatically updated online, and the demand state of the valve equipment is set or manually updated; finally, through the state change process F _i→j and the secondary update The corresponding relationship of the program update rule Φi _→j (X) (as shown in Table 5 and Table 6), and update the corresponding modular fault tree reliability model in time. Finally, when needed for each update or risk calculation, the fault tree model corresponding to each valve equipment in the nuclear power plant will obtain the corresponding structural function through collective calculation, and substitute each basic event into the basic reliability model adopted by the nuclear power plant to calculate The unreliability R(X,t) of valve body failure provides input for the nuclear power plant risk model.

Claims (1)

1. a kind of monitoring reliability method of nuclear power plant's valve body failure based on fault tree, which is characterized in that including as follows Step：

(1) status monitoring information of online acquisition nuclear power plant valve：

The monitoring signals that valve arrangement is acquired from the information monitoring of nuclear power plant and management system are specifically included from real time monitoring system The switching signal of system acquisition valve acquires valves test/failure logging signal, from Digital Control from auxiliary isolation computing system The startup of system acquisition valve controls signal；According to above-mentioned signal and information, the status monitoring sign of a valve arrangement is built The member that space Ω, Ω include is known as：The normally opened signal 1 of valve, the normal OFF signal 0 of valve, maintenance of valve/failure logging signal r, valve Reach pass step control signal -1, valve is closed to opening step control signal+1；Wherein, the normally opened signal of valve and normal OFF signal are referred to as For valve constant signal；Valve reaches Guan Heguan and is referred to as valve step signal to step control signal is opened；Valve arrangement state Monitoring sign space Ω is expressed as：

Ω=(1,0, r, -1 ,+1)；

(2) original state of online recognition valve and original state transfer figure is established；

(2.1)：Establish the initial state space S of valve arrangement_o；S_oIncluding state have：Open s₁, close s₂, failure s₃, preventive maintenance s₄, correct repair s₅；Valve arrangement initial state space S_oIt is expressed as：

S_O=(s₁,s₂,s₃,s₄,s₅)

By the status monitoring sign space Ω of valve arrangement, detects valve and be in state space S in actual operation_oIn Which kind of original state determines that valve monitors sign space Ω to valve arrangement initial state space S_oCriterion function f_C, method is such as Under：

State is to open when (2.1.1) valve constant signal is 1；

State is pass when (2.1.2) valve constant signal is 0；

(2.1.3) failure logging signal r determines that state is to correct repair, preventive maintenance or failure；

(2.1.4) valve step signal+1 or -1 represents the automatic startup of valve state on or off；

Establish following mapping relations：

(2.2) valve original state transfer figure is established according to actual motion；It is as follows that original state shifts figure method for building up：

The valve arrangement initial state space S of (2.2.1) based on foundation_OThe original state for determining valve, is indicated with oval frame；

(2.2.2) determines the transfer path between original state based on power plant's actual motion, with solid line table with the arrow and digital Show, transfer path serial number between digital representation original state；

Determining complete transfer path includes：In normal course of operation, s is opened₁With pass s₂Mutual conversion；In maintenance process, s is opened₁ With preventive maintenance s₃It is mutual conversion and pass s₂With preventive maintenance s₃Mutual conversion；In failure procedure, s is opened₁To failure s₃Turn Change and close s₂To failure s₃Conversion；In repair process, failure s₃S is repaired to correcting₅Conversion, correct repair s₅To opening s₁Turn Change and correct repair s₅To pass s₁Conversion；Valve arrangement status monitoring sign space Ω and criterion function f_CMonitor original state Or change, relevant monitoring signals input is added in supplement in state transition diagram；

(3) the modularization fault tree reliability model and online updating of the failure of valve arrangement ontology are established：

(3.1) original state-transferring demand situations figure of valve arrangement is established：

Establish the need state space S of valve arrangement_d, S_dIncluding state have：Open S_d1, close S_d2, no demand O；

S_d=(s_d1,s_d2,o)；

Valve arrangement original state is established to the mapping relations of need state, establishes rule f_sIt is as follows：

(3.1.1) " original state " chooses defined initial state space S₀In element；

(3.1.2) " need state " chooses defined need state space S_dIn element；

" need state " is sky when (3.1.3) valve arrangement " original state " is determined as unavailable；

When (3.1.4) valve arrangement " original state " is determined as available mapping object is determined by actual motion；

Rule-based f_s, the mapping relations of valve arrangement original state to need state are：Valve arrangement is initially out S₁State When, power plant is defined as can be used, and whole demands include opening S_d1With pass S_d2State；Valve, which is initially, closes S₁When state, power plant is defined as It can use, whole demands include opening S_d1With pass S_d2State；Valve arrangement is initially failure S₃, preventive maintenance S₄, correct repair S₅Shape When state, power plant is defined as unavailable, and need state is sky at this time, corresponds to that " mapping relations that no demand O is established are as follows：

The complete original state of valve arrangement-transferring demand situations figure is established on the basis of original state transfer figure, establishes rule For：

The valve arrangement need state space S of (3.1.1.1) based on foundation_dThe need state for determining valve, is indicated with box；

(3.1.1.2) based on power plant's actual motion determine original state to need state transfer path, with the arrow and alphabetical Solid line indicate, letter indicate original state to need state transfer path serial number；

It is regular according to establishing by the mapping relations of above-mentioned valve arrangement original state to need state, at the beginning of nuclear power plant's valve Original state-transferring demand situations figure of complete valve arrangement is established out in the monitoring of beginning state and transfer figure；

(3.2) the modularization fault tree reliability model of valve arrangement ontology failure is established：

Valve arrangement modularization failure tree modeling method is as follows：

(3.2.1) modularization fault tree does not define equipment original state；

(3.2.2) modularization fault tree envelope equipment state shifts the failure mode of all analysis states in figure；

After establishing out the modularization fault tree models of equipment, the minimal cut set set of fault tree is solved by Boolean calculation；Such as Fruit includes N number of minimal cut set, then j-th of minimal cut set is expressed as：

K_j=(X^j ₁,X^j ₂,…,X^j _i,…·,X^j _n)

X^j _iIndicate that failure mode, the minimal cut set set expression of modularization fault tree are：

Θ={ K₁,K₂,…,K_N}

In Θ, as long as any one minimal cut set K_jWhole bottom event X^j _iOccur, wherein j=1,2 ..., N, the top of fault tree Event must occur, and minimal cut set set is used for indicating the structure function of fault tree；Each fault tree synthesis function is to pass through After the set operation of fault tree simplify, absorb as a result, i.e. modularization fault tree synthesis function Θ₀(X) it is expressed as：

The failure mode that all kinds of valves include is as follows：

(3.2.1.1) motor-driven valve：Rupture, leakage, interior leakage are refused out, refuse to close, open by mistake, and accidentally close；

(3.2.1.2) check-valves：Rupture, leakage, interior leakage are refused out；

(3.2.1.3) hand-operated valve：Rupture, leakage, interior leakage are forgotten to open, forget to close；

(3.2.1.4) pneumatic operated valve：Rupture, leakage, interior leakage are accidentally closed, and are accidentally closed, manually opened failure；

When valve arrangement is in " preventive maintenance ", " correct repair ", " failure " original state, it is unavailable to represent its generation, at this time Only include the unavailable failure mode of failure/repair, carries out the unavailable logical renewal processing of fault tree；

(3.3) the update rule for establishing out valve arrangement modularization fault tree, obtains modularization fault tree to each status fault tree Mapping relations：

(3.3.1) is if by X in modularization fault tree^j _iEvent attribute is set as False, indicates that event determination does not occur；So In fault tree synthesis function Θ₀(X) remove comprising X in^j _iAll minimal cut sets, re-start set operation absorb, simplify after Generate new structure function Θ_i(X)；

(3.3.2) is if by X in modularization fault tree^j _iEvent attribute is set as True, indicates that the event has occurred；In fault tree knot Structure function Θ₀(X) remove the X for including in minimal cut set in^j _iElement re-starts set operation and absorbs, generated newly after simplification Structure function Θ_i(X)；

(3.3.3) is if by X in modularization fault tree^j _iEvent attribute is set as Normal, indicates that the event is occurred with certain probability, Then substitute into the calculating of elementary event reliability model；That is fault tree synthesis function Θ₀(X) do not change；

On the basis of valve module fault tree models, the fault tree models of each state are obtained；Pass through modularization failure Tree construction function Θ₀(X), practical mapping relationship f is established_x, obtain the structure function of each status fault tree-model of equipment；Mapping table It is shown as：

For each determining valve arrangement state, all corresponding there are one status fault tree construction function Θ_i(X)：

Θ_i(X)=f_x ⁱ[Θ₀(X)]

It is as follows to change the secondary update of newer characteristic for valve need state：

(3.3.3.1) once updates：It monitors that valve original state changes, if need state is identical as original state, carries out valve The update of door state；Monitor that valve original state is constant, power plant's configuration variation causes valve need state to update, and directly carries out The update of valve state；

(3.3.3.2) secondary update：For monitoring that valve original state changes situation in once updating, judging that valve is practical needs Ask state whether identical as original state, the update of identical then Exactly-once, different then it needs to be determined that variation need after original state Ask the secondary update of state；Establish the secondary update rule of modularization fault tree：

When each valve state updates, a F is defined_iState is to F_jThe transfer process of state：

F_i→j=F_j-F_i

F_i→jCorrespond to the transfer path in state transition diagram；In Θ (X), there are one F_iThe fault tree synthesis function Θ of state_i (X) to F_jThe fault tree synthesis function Θ of state_j(X) change procedure：

Θ_i→j(X)=Θ_j(X)-Θ_i(X)

The Φ of generation_i→j(X) it is exactly update rule, and has following correspondence：

F_i→j——→Φ_i→j(X)

In valve body failed module bottom event of fault tree, rupture, leakage can be summarized as ON/OFF operational failure；Interior leakage is to need It asks to close operational failure；It refuses out, refuse to close, opening by mistake, accidentally closing and can be summarized as demand expiration；Wherein refuse out and open by mistake is closed to the mistake opened Journey claims demand expiration 1；It is the process for reaching pass to refuse to close and accidentally close, and claims demand expiration 2；

Primary more new route：The path 1,2,3,4,5,6,7,8,9,10,11, a, d；

Secondary more new route：The path 1-d, 2-a, 6-a, 7-d, 9-a, 11-d；

(3.4) judge the state of valve and update rule is executed to its modularization fault tree；First determine whether valve " original state ", If constant judgement valve need state, if variation executes its modularization fault tree after updating for the first time judges valve demand shape again State；Then in the constant judgement valve need state of original state, its modularization fault tree needn't be executed if constant Update, if same execute of variation updates for the first time；Finally valve need state is judged in original state variation, if constant Second then needn't be executed to its modularization fault tree to update, updated if changing and also needing to execute second；

(4) reliability that the valve arrangement ontology based on fault tree fails in line computation；

The modularization fault tree for establishing valve body failure carries out modularization fault tree by the update rule of call establishment Logical assignment；By the Boolean set operations of fault tree models the minimal cut set set of reliability model is obtained after simplified, absorption； It is finally translated into corresponding fault tree synthesis function, corresponding elementary event reliability model is substituted into after carrying out non cross link processing, Calculate the failure probability of valve body failure；The structure function Φ (X) converted by minimal cut set set is：

Substitute into t moment elementary event X^j _iThe obtained elementary event failure probability P of reliability model^j _i(t), valve arrangement sheet is obtained The unreliable degree R (X, t) of body failure, i.e.,：