Monte Carlo reliability simulation method based on SysML activity diagram

Abstract

The invention provides a Monte Carlo reliability simulation method based on a SysML activity diagram, and relates to the technical field of system reliability modeling and reliability evaluation. The Monte Carlo reliability simulation method based on the SysML activity diagram comprises the following steps: step 1, carrying out system reliability modeling based on the SysML activity diagram; step 2, fault simulation based on discrete events; and step 3, reliability simulation based on Monte Carlo. According to the method, the SysML activity diagram is expanded, the reliability meta-model is built, then the system reliability block diagram model is built, a unified data source of system design and system reliability simulation is formed based on SysML, an information island between the system design and the system reliability simulation is broken through, the system design analysis efficiency is improved, the SysML is adopted to build the system reliability block diagram model, a tracing mechanism is easily built, and the system reliability simulation efficiency is improved. When the system design is changed, the system reliability block diagram can be correspondingly changed in time, and the consistency is ensured.

Description

Technical Field

[0001] The present invention relates to the field of system reliability modeling technology and reliability evaluation technology, in particular to a Monte Carlo reliability simulation method based on SysML activity diagram. Background Art

[0002] In recent years, with the rapid development of science and technology, digital, intelligent and information equipment has been widely used in various systems, resulting in the continuous expansion of system scale and increasingly complex system characteristics. The resulting maintenance costs have gradually increased, even exceeding 60% of the system's full life cycle costs. Therefore, improving system reliability has become an important issue that cannot be ignored in the system design and development process.

[0003] In order to analyze and evaluate the reliability of a system, it is usually necessary to establish a reliability model of the system. In engineering practice, reliability block diagram models are widely used to identify and analyze the reliability logical relationship of each component device in the system. As the development and implementation process of complex systems increasingly relies on modeling technology, the Model Based System Engineering (MBSE) method has emerged. MBSE uses SysML as the standard modeling language to support the description, analysis, design, verification and validation of complex systems. As an important method and means in the future development process of complex systems, MBSE also needs to be matched with technical methods in the professional dimensions of reliability analysis and design.

[0004] Currently, the nine SysML diagrams do not support the construction of a system reliability block diagram model. Considering that the reliability block diagram contains start, end, unit nodes, and logic nodes, which are similar to the SysML activity diagram, a reliability metamodel is built based on SysML activity diagram elements, and then a reliability block diagram model is built. This can effectively break through the information islands between system design and system reliability analysis, and improve the efficiency of system design and analysis.

[0005] In the field of system reliability analysis, most reliability models only describe the static characteristics of the system, but rarely describe the dynamic characteristics of the system. Even the existing reliability models based on Petri nets, Markov, etc. use strong assumptions and constraints to analyze dynamic models, so it is difficult to meet the needs of complex system reliability analysis.

[0006] Therefore, those skilled in the art provide a Monte Carlo reliability simulation method based on SysML activity diagram to solve the problems raised in the above background technology. Summary of the invention

[0007] 1. Technical issues to be resolved

[0008] In view of the deficiencies in the prior art, the present invention provides a Monte Carlo reliability simulation method based on SysML activity diagram. The method expands the SysML activity diagram to build a system reliability block diagram, breaks through information islands to describe the dynamic characteristics of the system, and then uses the Monte Carlo simulation method to implement system reliability analysis, which greatly improves the system's reliability analysis capability and solves the problems of information fragmentation in the process of system design and system reliability analysis and insufficient description of the system's dynamic characteristics by the current reliability analysis.

[0009] (II) Technical solution

[0010] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0011] A Monte Carlo reliability simulation method based on SysML activity diagram includes the following steps:

[0012] Step 1: System reliability modeling based on SysML activity diagram

[0013] First, the reliability metamodel is extended based on the SysML activity diagram, and then the system reliability block diagram model is built. The system reliability model can be composed of basic reliability models such as series connection, parallel connection, bypass connection and voting. Then, the failure distribution type and failure distribution parameters of each device are configured to complete the construction of the system reliability model.

[0014] Step 2: Fault simulation based on discrete events

[0015] First, random numbers are generated based on the failure distribution function of the equipment unit as equipment failure events, and they are sorted in the event table. Then, equipment failure events are extracted in sequence. If the equipment is not enabled, the equipment failure event is regenerated and added to the event table. If the equipment is enabled, the impact of its failure on the system is determined based on the reliability logic relationship. If a system failure is caused, the system failure time is recorded and the number of system failures is increased by 1. If no system failure is caused, the equipment failure event is regenerated and added to the event table, and equipment failure events are continuously extracted in sequence and failure simulation is performed.

[0016] Step 3: Monte Carlo-based reliability simulation

[0017] First, perform simulation configuration, set the number of simulations, simulation time, etc. If the number of simulations is less than the simulation setting number, call the discrete event-based fault simulation to perform system fault simulation, and count system fault information and fault number and other indicators. When the number of simulations reaches the simulation setting number, calculate the system reliability based on the MTBF indicator to obtain the system reliability curve.

[0018] Furthermore, in step one, SysML is a modeling language for system engineering. It provides a unified modeling method for complex systems involving multiple disciplines such as mechanics, electronics, control, and hydraulics, forms a unified data source, and ensures semantic consistency. At the same time, it forms a unified data source for system design and system reliability simulation, breaks through the information islands between the two, and improves the efficiency of system design and analysis.

[0019] Furthermore, in the step 2, the fault simulation based on discrete events adopts the next earliest event advancement method, which is different from the fixed increment advancement method, thereby ensuring that all equipment failure events are covered in the simulation process and improving the simulation efficiency.

[0020] Furthermore, in the step three, Monte Carlo reliability simulation is performed based on the SysML reliability model and discrete event failure simulation, which is suitable for reliability simulation of complex systems with large scale and strong systematicity, and improves the reliability simulation capability of complex systems.

[0021] Furthermore, in the step 1, the system reliability block diagram model is divided into two parts: RBD node metamodel construction and logic node metamodel construction:

[0022] The RBD node metamodel is extended based on the Act ion element to obtain the element template of the RBD node. Reliability analysis based on the system reliability block diagram requires input of information such as equipment failure distribution type and failure distribution parameter, so its attributes are extended: failure distribution type, MTBF and failure distribution parameter 2. Among them, the failure distribution type is Boolean type, including normal distribution, exponential distribution, Weibull distribution and binomial distribution, and MTBF and failure distribution parameter 2 are real number type;

[0023] The logical node metamodel is expanded based on the Dec is ionNode element to obtain the bypass node. The backup priority needs to be considered when the bypass connection is made, so its attribute is extended: priority sorting, the type is string type; based on the MergeNode element, it is expanded to obtain the voting node. When the voting connection is made, the number of faulty devices needs to be considered to cause the overall failure, so its attribute is extended: r, the number of faulty devices, the type is real number type.

[0024] Furthermore, the system reliability block diagram model is composed of basic reliability meta-models such as series connection, parallel connection, bypass connection and voting, among which bypass connection and voting require bypass connection logic nodes and voting logic nodes. Based on the above reliability meta-model, the system reliability model can be built. First, its basic reliability model is introduced as follows:

[0025] In the serial model, the RBD nodes can be connected in sequence using control flow. When any device is damaged, the model cannot run. That is, the failure of any unit in all the units that make up the system will cause the failure of the entire system. Assume that the system model consists of n units and the state of the i-th unit is S i , when S i =1 (i=1,2,3,…,n), the system state S=1, the system works normally;

[0026] Parallel model, add a vertical branch node before the RBD node, and add a merge node after the RBD node, which means that when the system is running, all devices are running and the entire model can run as long as one device is running. That is, when all units that make up the system fail, the system will fail and become a hot standby system. Assume that the system model consists of n units and the state of the i-th unit is S i , when S i =0 (i=1,2,3,…,n), the system state S=0, the system fails.

[0027] In the bypass model, add a bypass node before the RBD node, and add a merge node after the RBD node in the same way as in the parallel connection. Set the startup order of each device in the priority attribute of the bypass node, which means that when the system is running, the devices participate in the operation in turn, that is, only one of the system units participates in the work at the same time. When the participating unit fails, another unit is enabled to work through the conversion device. When all the units that make up the system fail, the system will fail. It is also called a cold backup system.

[0028] Due to its special working mode, if the failure of the conversion device is not considered, the system state can be directly judged by the state of the last unit. Assuming that the system model consists of n units, the state of the nth unit is S n , when S i =1 (i=1,2,3,…,n), the system state S=1, the system works normally;

[0029] Voting model, add a vertical branch node before the RBD node, add a voting node after the RBD node and set the parameter r, which means that when the system is running, all devices are running and the number of normal units in the n units that make up the system is not less than r, the system will not fail. Such a system is called a voting model. Assume that the system model consists of n units, and the state of the i-th unit is S i , let x be the number of units required for the normal operation of the system, y be the number of units working normally in the current system, when y>x, the system works normally.

[0030] Furthermore, in the step 2, the discrete event-based fault simulation process mainly includes the following steps:

[0031] Step 2.1: Single simulation initialization: clear the fault event table;

[0032] Step 2.2: Generate failure events: traverse all equipment units in the reliability model, and generate random numbers based on the failure distribution function of the equipment units as equipment failure events;

[0033] Step 2.3: Save the equipment fault information into the event table;

[0034] Step 2.4: Sort the equipment failure events in the event table according to the time sequence of the failure events;

[0035] Step 2.5: Determine whether there is a fault event in the fault event table. If not, the simulation ends. If so, go to step 2.6.

[0036] Step 2.6: According to the fault event sorting results, read the fault event that currently ranks first;

[0037] Step 2.7: Determine whether the device is in an enabled state. If the device is in an enabled state, proceed to step 2.8; if not, proceed to step 2.12;

[0038] Step 2.8: Determine whether the device is a series model, if so, proceed to step 2.9, if not, proceed to step 2.10;

[0039] Step 2.9: The device failure causes the system failure, the system failure time is recorded and the number of system failures increases by 1;

[0040] Step 2.10: If it is not a series model, it is necessary to determine whether the basic reliability model composed of the device is invalid based on its connection type and whether its backup device is damaged;

[0041] Step 2.11: If it fails, go to step 2.9; if it does not fail, go to step 2.12;

[0042] Step 2.12: According to the reliability distribution function of the equipment, re-extract the equipment failure events and proceed to step 2.3.

[0043] Furthermore, in step 3, the Monte Carlo-based reliability simulation process mainly includes the following steps:

[0044] Step 3.1: Set simulation conditions: simulation duration, number of simulations, etc.

[0045] Step 3.2: Simulation times i = 1;

[0046] Step 3.3: Determine whether the number of simulations is greater than the number of simulation settings. If so, proceed to step 3.4. If less than or equal to, proceed to step 3.5.

[0047] Step 3.4: Calculate the system reliability based on the statistical results, and then end;

[0048] Step 3.5: Call step 2 to perform fault simulation based on discrete event simulation principle;

[0049] Step 3.6: According to the discrete event fault simulation results, statistics on system fault information and fault times and other indicators;

[0050] Step 3.7: Increase the number of simulations by 1, and then go to step 3.3.

[0051] (III) Beneficial effects

[0052] The present invention provides a Monte Carlo reliability simulation method based on SysML activity diagram, which has the following beneficial effects:

[0053] 1. The present invention provides a Monte Carlo reliability simulation method based on SysML activity diagram. The method expands the SysML activity diagram, builds a reliability metamodel, and then builds a system reliability block diagram model. Based on SysML, a unified data source for system design and system reliability simulation is formed, which connects the information islands between the two and improves the efficiency of system design and analysis.

[0054] 2. The present invention provides a Monte Carlo reliability simulation method based on SysML activity diagram. The method uses SysML to build a system reliability block diagram model, which is conducive to establishing a traceability mechanism. When the system design is changed, the system reliability block diagram can be changed accordingly in time to ensure consistency.

[0055] 3. The present invention provides a Monte Carlo reliability simulation method based on SysML activity diagram. The method is based on the Monte Carlo reliability simulation of discrete event mechanism, which improves the efficiency of system reliability analysis while considering the dynamic characteristics of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A Monte Carlo reliability simulation flow chart based on SysML activity diagram of the present invention;

[0057] Figure 2 It is a composition diagram of the RBD node metamodel of the present invention;

[0058] Figure 3 It is a composition diagram of the logical node metamodel of the present invention;

[0059] Figure 4 It is a composition diagram of the SysML series group model of the present invention;

[0060] Figure 5 It is a composition diagram of the SysML parallel group model of the present invention;

[0061] Figure 6 It is a composition diagram of the SysML side-link group model of the present invention;

[0062] Figure 7 An interface diagram for setting the priority sorting of side links of the present invention;

[0063] Figure 8 It is a composition diagram of the SysML voting group model of the present invention;

[0064] Fig. 9 The voting r value setting interface diagram of the present invention;

[0065] Fig.10 A discrete event-based fault simulation flow chart of the present invention;

[0066] Fig.11 The reliability simulation flow chart based on Monte Carlo of the present invention;

[0067] Fig.12 It is a composition diagram of the SysML system reliability model of the present invention;

[0068] Fig.13 is a fault event table of the present invention;

[0069] Fig.14 It is a system reliability curve diagram of the present invention. DETAILED DESCRIPTION

[0070] The following will be combined with the drawings in the specific embodiments of the present invention to clearly and completely describe the technical solutions in the specific embodiments of the present invention. Obviously, the specific embodiments described are only part of the specific embodiments of the present invention, not all of the specific embodiments. Based on the specific embodiments of the present invention, all other specific embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Specific implementation method:

[0072] The overall process is as follows Figure 1 As shown, a specific embodiment of the present invention provides a Monte Carlo reliability simulation method based on SysML activity diagram, comprising the following steps:

[0073] Step 1: System reliability modeling based on SysML activity diagram

[0074] First, the reliability metamodel is extended based on the SysML activity diagram, and then the system reliability block diagram model is built. The system reliability model can be composed of basic reliability models such as series connection, parallel connection, bypass connection and voting. Then, the failure distribution type and failure distribution parameters of each device are configured to complete the construction of the system reliability model.

[0075] The system reliability block diagram model is divided into two parts: RBD node metamodel construction and logical node metamodel construction:

[0076] RBD node metamodel, such as Figure 2 As shown in the figure, based on the Act i on element, the element stereotype of the RBD node is obtained. Reliability analysis based on the system reliability block diagram requires input of equipment failure distribution type, failure distribution parameter and other information, so its attributes are extended: failure distribution type, MTBF and failure distribution parameter 2, among which, the failure distribution type is Boolean type, including normal distribution, exponential distribution, Weibull distribution and binomial distribution, and MTBF and failure distribution parameter 2 are real number type;

[0077] Logical node metamodel, such as Figure 3 As shown in the figure, based on the Dec isi onNode element, the bypass node is obtained. When the bypass connection is made, the backup priority needs to be considered, so its attribute is extended: priority sorting, the type is string type; based on the MergeNode element, the voting node is obtained. When the voting connection is made, the number of faulty devices needs to be considered to cause the overall failure, so its attribute is extended: r, the number of faulty devices, the type is real number type;

[0078] The system reliability block diagram model is composed of basic reliability meta-models such as series connection, parallel connection, bypass connection and voting. The bypass connection and voting require bypass logic nodes and voting logic nodes. The system reliability model can be built based on the above reliability meta-model. First, its basic reliability model is introduced as follows:

[0079] Series models, such as Figure 4 As shown, the RBD nodes can be connected in sequence using control flow. When any device is damaged, the model cannot run, that is, the failure of any unit in all the units that make up the system will cause the failure of the entire system. Assume that the system model consists of n units and the state of the i-th unit is S i , when S i =1 (i=1,2,3,…,n), the system state S=1, the system works normally;

[0080] Parallel models, such as Figure 5 As shown in the figure, a vertical branch node (Fork Vertical) is added before the RBD node, and a merge node (Merge) is added after the RBD node, which means that when the system is running, all devices are running and the entire model can run as long as one device is running. That is, when all units that make up the system fail, the system will fail and become a hot standby system. Assume that the system model consists of n units and the state of the i-th unit is S i , when Si =0 (i=1,2,3,…,n), the system state S=0, the system fails;

[0081] Parallel models, such as Figure 6 As shown, add a bypass node before the RBD node, add a merge node (Merge) after the RBD node in the same way as the parallel connection, and set the startup order of each device in the priority attribute of the bypass node, as shown in Figure 7 As shown, it means that when the system is running, the devices participate in the operation in sequence, that is, only one of the system units participates in the work at the same time. When the participating unit fails, another unit is enabled to work through the conversion device. When all the units that make up the system fail, the system will fail. It is also called a cold backup system.

[0082] Due to its special working mode, if the failure of the conversion device is not considered, the system state can be directly judged by the state of the last unit. Assuming that the system model consists of n units, the state of the nth unit is S n , when S i =1 (i=1,2,3,…,n), the system state S=1, the system works normally;

[0083] Voting models, such as Figure 8 As shown, add a vertical branch node (Fork Vertical) before the RBD node, add a voting node after the RBD node and set the parameter r, as shown in Fig. 9 As shown, it means that when the system is running, all devices are running and the number of normal units among the n units that make up the system is not less than r (1≤r≤n), the system will not fail. Such a system is called a voting model. Assume that the system model consists of n units, and the state of the i-th unit is S i , let x be the number of units required for the normal operation of the system, y be the number of units that are currently operating normally in the system, when y>x, the system is operating normally;

[0084] Step 2: Fault simulation based on discrete events

[0085] First, random numbers are generated based on the failure distribution function of the equipment unit as equipment failure events, and they are sorted in the event table. Then, equipment failure events are extracted in sequence. If the equipment is not enabled, the equipment failure event is regenerated and added to the event table. If the equipment is enabled, the impact of its failure on the system is determined based on the reliability logic relationship. If a system failure is caused, the system failure time is recorded and the number of system failures is increased by 1. If no system failure is caused, the equipment failure event is regenerated and added to the event table, and equipment failure events are continuously extracted in sequence and failure simulation is performed.

[0086] The discrete event-based fault simulation process mainly includes the following steps:

[0087] Step 2.1: Single simulation initialization: clear the fault event table;

[0088] Step 2.2: Generate failure events: traverse all equipment units in the reliability model, and generate random numbers based on the failure distribution function of the equipment units as equipment failure events;

[0089] Step 2.3: Save the equipment fault information into the event table;

[0090] Step 2.4: Sort the equipment failure events in the event table according to the time sequence of the failure events;

[0091] Step 2.5: Determine whether there is a fault event in the fault event table. If not, the simulation ends. If so, go to step 2.6.

[0092] Step 2.6: According to the fault event sorting results, read the fault event that currently ranks first;

[0093] Step 2.7: Determine whether the device is in an enabled state. If the device is in an enabled state, proceed to step 2.8. If it is not enabled (in a standby state), proceed to step 2.12.

[0094] Step 2.8: Determine whether the device is a series model, if so, proceed to step 2.9, if not, proceed to step 2.10;

[0095] Step 2.9: The device failure causes the system failure, the system failure time is recorded and the number of system failures increases by 1;

[0096] Step 2.10: If it is not a series model, it is necessary to determine whether the basic reliability model composed of the equipment is invalid based on its connection type (parallel, bypass, voting) and whether its backup (cold or hot backup) equipment is damaged;

[0097] Step 2.11: If it fails, go to step 2.9; if it does not fail, go to step 2.12;

[0098] Step 2.12: According to the reliability distribution function of the equipment, re-extract the equipment failure event and go to step 2.3;

[0099] Step 3: Monte Carlo-based reliability simulation

[0100] First, perform simulation configuration, set the number of simulations, simulation time, etc. If the number of simulations is less than the simulation setting number, call the discrete event-based fault simulation to perform system fault simulation, and count system fault information and fault number and other indicators. When the number of simulations reaches the simulation setting number, calculate the system reliability based on the MTBF indicator to obtain the system reliability curve.

[0101] The Monte Carlo method is a statistical test method based on random numbers to solve computational problems. The Monte Carlo method is used to simulate and analyze the system reliability model. The Monte Carlo-based reliability simulation process mainly includes the following steps:

[0102] Step 3.1: Set simulation conditions: simulation duration, number of simulations, etc.

[0103] Step 3.2: Simulation times i = 1;

[0104] Step 3.3: Determine whether the number of simulations is greater than the number of simulation settings. If so, proceed to step 3.4. If less than or equal to, proceed to step 3.5.

[0105] Step 3.4: Calculate the system reliability based on the statistical results, and then end;

[0106] Step 3.5: Call step 2 to perform fault simulation based on discrete event simulation principle;

[0107] Step 3.6: According to the discrete event fault simulation results, statistics on system fault information and fault times and other indicators;

[0108] Step 3.7: Add 1 to the number of simulations, and then go to step 3.3;

[0109] in,

[0110] MTBF = system operation time / number of system failures;

[0111] According to MTBF, the reliability of the system can be obtained by substituting it into the formula. The calculation formula is as follows:

[0112] R(t)=e (-t / MTBF)

[0113] In the step 1, SysML is a modeling language for system engineering. It provides a unified modeling method for complex systems involving multiple disciplines such as mechanics, electronics, control, and hydraulics, forms a unified data source, and ensures semantic consistency. At the same time, it forms a unified data source for system design and system reliability simulation, breaks through the information islands between the two, and improves the efficiency of system design and analysis.

[0114] In the step 2, the fault simulation based on discrete events adopts the next earliest event advancement method, which is different from the fixed increment advancement method. This ensures that all equipment failure events are covered in the simulation process and improves the simulation efficiency.

[0115] In the step three, Monte Carlo reliability simulation is performed based on the SysML reliability model and discrete event failure simulation, which is suitable for reliability simulation of complex systems with large scale and strong systematicity, and improves the reliability simulation capability of complex systems.

[0116] The following is a practical example:

[0117] Build a system simulation case, such as Fig.12 As shown, device 2 and device 3 form a parallel group model, which is called subsystem 1;

[0118] Device 4 and device 5 form a bypass group model, and device 5 is the backup of device 4, which is called subsystem 2;

[0119] Subsystem 1 and subsystem 2 form a parallel group, which is called subsystem 1;

[0120] Equipment 1 and subsystem 1 form a series group;

[0121] The failure distribution type is set to exponential distribution, the mean time between failures MTBF = 300, and the bypass component setting form is the same as Figure 7 The same as in the previous section, so no further description is given.

[0122] Set the number of simulations to 1000 times, the simulation time to 1000 hours, and start the simulation when the current number of simulations i=1;

[0123] The current simulation number i is less than 1000, so discrete event simulation is performed.

[0124] According to the failure distribution function of each device, the failure events of each device are extracted and sorted in the event table. The failure event table is as follows: Fig.13 shown.

[0125] The first event is extracted. At 85h, device 2 fails, but does not cause a system failure. Therefore, after extracting the fault 2 event and arranging it in the event table, the next event in the fault event table is extracted, and the system fails. Taking this fault table as an example, after the system runs for 448h, the failure of device 1 causes a system failure. Therefore, the simulation ends, the system failure time is recorded, and the number of system failures is increased by 1, ending the first round of simulation.

[0126] Repeat the above steps for the number of simulations i+1 until i=1000, which means the number of simulations is reached. Calculate the system MTBF based on the statistical system failure information and the number of system failures. Finally, calculate the system reliability based on the MTBF. The reliability curve is as follows: Fig.14 shown.

[0127] The present specification uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

[0128] Although specific embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the specific embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A Monte Carlo reliability simulation method based on SysML activity diagram, characterized in that: The following steps are involved: Step 1: System reliability modeling based on SysML activity diagram First, the reliability metamodel is extended based on the SysML activity diagram, and then the system reliability block diagram model is built. The system reliability model can be composed of basic reliability models such as series connection, parallel connection, bypass connection and voting. Then, the failure distribution type and failure distribution parameters of each device are configured to complete the construction of the system reliability model. Step 2: Fault simulation based on discrete events First, random numbers are generated based on the failure distribution function of the equipment unit as equipment failure events, and they are sorted in the event table. Then, equipment failure events are extracted in sequence. If the equipment is not enabled, the equipment failure event is regenerated and added to the event table. If the equipment is enabled, the impact of its failure on the system is determined based on the reliability logic relationship. If a system failure is caused, the system failure time is recorded and the number of system failures is increased by 1. If no system failure is caused, the equipment failure event is regenerated and added to the event table, and equipment failure events are continuously extracted in sequence and failure simulation is performed. Step 3: Monte Carlo-based reliability simulation First, perform simulation configuration, set the number of simulations, simulation time, etc. If the number of simulations is less than the simulation setting number, call the discrete event-based fault simulation to perform system fault simulation, and count system fault information and fault number and other indicators. When the number of simulations reaches the simulation setting number, calculate the system reliability based on the MTBF indicator to obtain the system reliability curve.

2. The Monte Carlo reliability simulation method based on SysML activity diagram according to claim 1, characterized in that: In the step 1, SysML is a modeling language for system engineering. It provides a unified modeling method for complex systems involving multiple disciplines such as mechanics, electronics, control, and hydraulics, forms a unified data source, and ensures semantic consistency. At the same time, it forms a unified data source for system design and system reliability simulation, breaks through the information islands between the two, and improves the efficiency of system design and analysis.

3. The Monte Carlo reliability simulation method based on SysML activity diagram according to claim 1, characterized in that: In the step 2, the discrete event-based fault simulation uses the next earliest event advancement method, which is different from the fixed increment advancement method. This ensures that all equipment failure events are covered during the simulation process and improves the simulation efficiency.

4. The Monte Carlo reliability simulation method based on SysML activity diagram according to claim 1, characterized in that: In the step three, Monte Carlo reliability simulation is performed based on the SysML reliability model and discrete event failure simulation, which is suitable for reliability simulation of complex systems with large scale and strong systematicity, and improves the reliability simulation capability of complex systems.

5. The Monte Carlo reliability simulation method based on SysML activity diagram according to claim 1, characterized in that: In the step 1, the system reliability block diagram model is divided into two parts: RBD node metamodel construction and logical node metamodel construction: The RBD node metamodel is extended based on the Action element to obtain the element template of the RBD node. Reliability analysis based on the system reliability block diagram requires input of information such as equipment failure distribution type and failure distribution parameter, so its attributes are extended: failure distribution type, MTBF and failure distribution parameter 2. Among them, the failure distribution type is Boolean type, including normal distribution, exponential distribution, Weibull distribution and binomial distribution, and MTBF and failure distribution parameter 2 are real number type; The logical node metamodel is extended based on the DecisionNode element to obtain the bypass node. When the bypass connection is made, the backup priority needs to be considered, so its attribute is extended: priority sorting, the type is string type; based on the MergeNode element, it is extended to obtain the voting node. When the voting connection is made, the number of faulty devices needs to be considered to cause the overall failure, so its attribute is extended: r, the number of faulty devices, the type is real number type.

6. The Monte Carlo reliability simulation method based on SysML activity diagram according to claim 5, characterized in that: The system reliability block diagram model is composed of basic reliability meta-models such as series connection, parallel connection, bypass connection and voting, among which bypass connection and voting require bypass connection logic nodes and voting logic nodes. Based on the above reliability meta-model, the system reliability model can be built. First, its basic reliability model is introduced as follows: In the serial model, the RBD nodes can be connected in sequence using control flow. When any device is damaged, the model cannot run. That is, the failure of any unit in all the units that make up the system will cause the failure of the entire system. Assume that the system model consists of n units and the state of the i-th unit is S i , when S i =1 (i=1,2,3,…,n), the system state S=1, the system works normally; Parallel model, add a vertical branch node before the RBD node, and add a merge node after the RBD node, which means that when the system is running, all devices are running and the entire model can run as long as one device is running. That is, when all units that make up the system fail, the system will fail and become a hot standby system. Assume that the system model consists of n units and the state of the i-th unit is S i , when S i =0 (i=1,2,3,…,n), the system state S=0, the system fails. In the bypass model, add a bypass node before the RBD node, and add a merge node after the RBD node in the same way as in the parallel connection. Set the startup order of each device in the priority attribute of the bypass node, which means that when the system is running, the devices participate in the operation in turn, that is, only one of the system units participates in the work at the same time. When the participating unit fails, another unit is enabled to work through the conversion device. When all the units that make up the system fail, the system will fail. It is also called a cold backup system. Due to its special working mode, if the failure of the conversion device is not considered, the system state can be directly judged by the state of the last unit. Assuming that the system model consists of n units, the state of the nth unit is S n , when S i =1 (i=1,2,3,…,n), the system state S=1, the system works normally; Voting model, add a vertical branch node before the RBD node, add a voting node after the RBD node and set the parameter r, which means that when the system is running, all devices are running and the number of normal units in the n units that make up the system is not less than r, the system will not fail. Such a system is called a voting model. Assume that the system model consists of n units, and the state of the i-th unit is S i , let x be the number of units required for the normal operation of the system, y be the number of units working normally in the current system, when y>x, the system works normally.

7. The Monte Carlo reliability simulation method based on SysML activity diagram according to claim 1, characterized in that: In the step 2, the discrete event-based fault simulation process mainly includes the following steps: Step 2.1: Single simulation initialization: clear the fault event table; Step 2.2: Generate failure events: traverse all equipment units in the reliability model, and generate random numbers based on the failure distribution function of the equipment units as equipment failure events; Step 2.3: Save the equipment fault information into the event table; Step 2.4: Sort the equipment failure events in the event table according to the time sequence of the failure events; Step 2.5: Determine whether there is a fault event in the fault event table. If not, the simulation ends. If so, go to step 2.

6. Step 2.6: According to the fault event sorting results, read the fault event that currently ranks first; Step 2.7: Determine whether the device is in an enabled state. If the device is in an enabled state, proceed to step 2.8; if not, proceed to step 2.12; Step 2.8: Determine whether the device is a series model, if so, proceed to step 2.9, if not, proceed to step 2.10; Step 2.9: The device failure causes the system failure, the system failure time is recorded and the number of system failures increases by 1; Step 2.10: If it is not a series model, it is necessary to determine whether the basic reliability model composed of the device is invalid based on its connection type and whether its backup device is damaged; Step 2.11: If it fails, go to step 2.9; if it does not fail, go to step 2.12; Step 2.12: According to the reliability distribution function of the equipment, re-extract the equipment failure events and proceed to step 2.

3.

8. The Monte Carlo reliability simulation method based on SysML activity diagram according to claim 1, characterized in that: In step 3, the Monte Carlo-based reliability simulation process mainly includes the following steps: Step 3.1: Set simulation conditions: simulation duration, number of simulations, etc. Step 3.2: Simulation number i = 1; Step 3.3: Determine whether the number of simulations is greater than the number of simulation settings. If so, proceed to step 3.

4. If less than or equal to, proceed to step 3.

5. Step 3.4: Calculate the system reliability based on the statistical results, and then end; Step 3.5: Call step 2 to perform fault simulation based on discrete event simulation principle; Step 3.6: According to the discrete event fault simulation results, statistics on system fault information and fault times and other indicators; Step 3.7: Increase the number of simulations by 1, and then go to step 3.3.

Images