Evaluation Methods for Common Cause Failures in Digital Instrumentation and Control Systems of Nuclear Power Plants

Abstract

A method for evaluating common-cause failures of software in a digital instrumentation and control system of a nuclear power plant is disclosed. The instrumentation and control system includes at least two instrumentation and control platforms. The evaluation method includes the following steps: obtaining modeling input; calculating the probability of basic events used for modeling based on the modeling input; establishing a fault tree, which includes basic events; and using the fault tree and the probability of basic events to analyze the failure probability of the digital instrumentation and control system. For any instrumentation and control platform, the basic events include: inter-platform common-cause failure basic events, which are basic events of common-cause failures occurring between this instrumentation and control platform and other instrumentation and control platforms; and intra-platform common-cause failure basic events, which are basic events of common-cause failures occurring within this instrumentation and control platform. Inter-platform common-cause failure basic events and intra-platform common-cause failure basic events are mutually exclusive events.

Description

Technical Field

[0001] This invention relates to the field of probabilistic safety assessment technology for nuclear power plants, and in particular to an evaluation method for common cause failures of digital instrumentation and control system software in nuclear power plants. Background Technology

[0002] The Digital Instrumentation and Control System (DCS) is the control center of a digital nuclear power plant, undertaking the core tasks of reactor control, protection, and monitoring. It specifically includes key functional modules such as the reactor protection system, diverse drive systems, and the plant control system, which are implemented using a combination of software and hardware. Compared to traditional analog systems, the operation of a digital instrumentation and control system is highly software-dependent; therefore, software-related failures are the main problems affecting the stable operation of the system.

[0003] Software failures typically exhibit systemic characteristics. A single software defect can cause the instrumentation and control platform system software or multiple functional application software to fail simultaneously, thus forming a common cause failure (CCF). This failure mode undermines the independence of redundant design in the instrumentation and control system, posing significant hidden dangers to the operation of the instrumentation and control system and even the entire nuclear power plant. Therefore, it is a critical issue that cannot be ignored in the safety assessment of nuclear power plants.

[0004] Currently, safety assessments of nuclear power plants primarily employ Probabilistic Safety Assessment (PSA), a risk quantification technique based on probability theory and statistics. It uses event tree and fault tree models to quantitatively analyze the overall safety of nuclear power plants. Given the importance of software common-cause failures in instrumentation and control system failures, these failures are a key focus in PSA analysis. However, current analytical methods for software common-cause failures are not mature enough, and the analytical models are often quite complex, or rely solely on expert judgment to simply assign values ​​to the impact of software common-cause failures. Furthermore, there is currently no universally applicable reliability data related to software common-cause failures within the industry.

[0005] Therefore, there is an urgent need to establish an evaluation method for common-cause failures of digital instrumentation and control systems in nuclear power plants, so as to facilitate reliable analysis and evaluation of common-cause failures in software. Summary of the Invention

[0006] This invention was made in view of the above-mentioned problems, and its purpose is to provide an evaluation method for common cause failures of software in the digital instrumentation and control system of nuclear power plants, so as to provide model-based and quantitative support for software common cause failure analysis of PSA in nuclear power plants.

[0007] To achieve the above objectives, this invention provides a method for evaluating common-cause failures of software in a digital instrumentation and control system of a nuclear power plant. The instrumentation and control system includes at least two instrumentation and control platforms. The evaluation method includes the following steps: obtaining modeling input; calculating the probability of basic events for modeling based on the modeling input; establishing a fault tree, the fault tree including the basic events; and using the fault tree and the probabilities of the basic events to analyze the failure probability of the digital instrumentation and control system. For any instrumentation and control platform, the basic events include: inter-platform common-cause failure basic events, which are basic events of common-cause failures occurring between this instrumentation and control platforms; and intra-platform common-cause failure basic events, which are basic events of common-cause failures occurring within this instrumentation and control platform. Furthermore, the inter-platform common-cause failure basic events and the intra-platform common-cause failure basic events are mutually exclusive events.

[0008] According to the evaluation method of this invention, a complete evaluation process for common-cause software failures is constructed through four steps: obtaining modeling input, calculating probabilities, building fault trees, and analyzing failure probabilities. This solves the problem of lack of systematicity and standardization in existing technologies, and provides a logically clear and traceable evaluation method for common-cause software failures in nuclear power plant PSAs. Furthermore, by classifying the basic events of common-cause software failures into two mutually exclusive categories—intra-platform and inter-platform—the risk of double-counting in quantitative fault tree analysis is avoided, while the fault tree modeling process is simplified.

[0009] Furthermore, in the evaluation method of the present invention, preferably, calculating the probability of the basic events used for modeling includes calculating the total probability of common-cause failure of the instrumentation and control platform software. For any instrumentation and control platform, the total probability of common-cause failure of the instrumentation and control platform software is equal to the sum of the probability of the basic event of common-cause failure of software between instrumentation and control platforms and the probability of the basic event of common-cause failure of software within the instrumentation and control platform. The total probability of common-cause failure of the instrumentation and control platform software is calculated using the following formula: U = (n · P 1 · V1 · V2 + n · P2 · V3 ) · P3 · In formula P4, U is the total probability of common cause failure of the instrumentation and control platform software, n is the number of lines of code in the software module, P1 is the probability of an error occurring in each line of code, V1 is the probability of no code error being found during independent analysis, V2 is the probability of no code error being found when the computer tests the high-level language program, V3 is the probability of no code error being found when the computer tests the machine language program, P2 is the probability of an error occurring in each line of code due to the compiler, P3 is the probability that the code error will cause the entire system to completely collapse, and P4 is the probability of a signal input that triggers a specific error code under a specific operating condition and causes common cause failure of the software.

[0010] Therefore, by comprehensively considering measurable or estimable engineering parameters such as the number of lines of code, the probability of errors per line of code, the test miss rate, the probability of compiler errors, the probability of system crashes, and the probability of triggering specific operating conditions, the probability of basic events can be calculated, which improves the reliability of quantitative results compared with the previous subjective assignment method.

[0011] Furthermore, in the evaluation method of the present invention, preferably, calculating the probability of the basic event used for modeling includes calculating the probability of the basic event of software common cause failure between the instrumentation and control platforms using the following formula: U1 = π · U, where U1 is the probability of the basic event of software common cause failure between the instrumentation and control platforms, and π is the proportion of the probability of the basic event of software common cause failure between the instrumentation and control platforms to the total probability of software common cause failure of the instrumentation and control platforms. Specifically, for any instrumentation and control platform, when the technical method of the instrumentation and control platform is different from that of other instrumentation and control platforms, π is 0; when the technical method of the instrumentation and control platform is the same as that of other instrumentation and control platforms but the architecture is different, π is 0.05; and when the technical method and architecture of the instrumentation and control platform are the same as those of other instrumentation and control platforms, π is 1.

[0012] Therefore, the share factor π enables a simple allocation of the probability of common software failures between platforms, and provides a discrete value rule for the share factor π based on the differences in technical methods or architectures between platforms. Thus, engineers can quickly and consistently determine the parameters according to the actual platform configuration without complex calculations, which facilitates the standardized implementation of the PSA model.

[0013] Furthermore, in the evaluation method of the present invention, preferably, the step of establishing a fault tree includes: taking the failure of the instrumentation and control system as the top event; making the top event include at least two branches: instrumentation and control system software failure and instrumentation and control system hardware failure; and adding the basic event of common software failure between instrumentation and control platforms and the basic event of common software failure within the instrumentation and control platform to the OR gate branch of the instrumentation and control system software failure.

[0014] Therefore, it is possible to simplify fault tree modeling for software common cause failures under instrumentation and control system failures, reduce modeling time and modeling scale, and improve the efficiency of analyzing software common cause failures and instrumentation and control system failures in nuclear power plant PSA.

[0015] Furthermore, in the evaluation method of the present invention, preferably, the modeling input includes the test data of the instrumentation and control system, and the parameters in calculating the total probability of common cause failure of the instrumentation and control platform software are obtained from the test data.

[0016] Therefore, data such as test reports and defect records available to the nuclear power plant itself can be used to assign parameter values, ensuring the accuracy and reliability of the evaluation results for the nuclear power plant.

[0017] The present invention also provides an electronic device, including a processor and a memory for storing instructions executable by the processor, the processor being used to execute the instructions to implement the above-described method for evaluating common-cause failures of digital instrumentation and control system software in nuclear power plants.

[0018] The present invention also provides a computer storage medium storing computer program code, which, when executed by a processor, implements the aforementioned method for evaluating common-cause failures of digital instrumentation and control system software in nuclear power plants.

[0019] The present invention also provides a computer program product, including computer program code, wherein when the computer program code is executed by a processor, the processor implements the steps in the above-described evaluation method for common cause failure of digital instrumentation and control system software in nuclear power plants. Attached Figure Description

[0020] Figure 1 This is a flowchart of the evaluation method for common cause failure of digital instrumentation and control system software in nuclear power plants according to the present invention.

[0021] Figure 2 This is an example of a constructed fault tree model. Detailed Implementation

[0022] The following description, in conjunction with the accompanying drawings, describes an embodiment of the present invention for evaluating common-cause failures in the digital instrumentation and control system software of a nuclear power plant (hereinafter referred to as the "evaluation method").

[0023] Figure 1 A flowchart of the evaluation method of this embodiment is shown.

[0024] In step S1, the modeling input is obtained. The modeling input obtained in step S1 includes the instrumentation and control system specification and test data. The instrumentation and control system specification includes numerical and non-numerical data such as the hardware architecture, software logic, and performance parameters of the digital instrumentation and control system. The instrumentation and control system test data includes test reports and defect records from the system's development, integration, and verification phases.

[0025] Next, in step S2, the values ​​of basic events are calculated. Specifically, based on the modeling input obtained in step S1, the probabilities of the basic events used for fault tree modeling are calculated. The basic events referred to here are the base events of the fault tree, that is, the smallest indivisible unit in the fault tree, and the probability of a basic event is the probability of its occurrence.

[0026] In this embodiment, the basic events calculated in step S2 include two basic events: the basic event of common software failure between instrumentation and control platforms and the basic event of common software failure within the instrumentation and control platform. The following explains the method for determining the values ​​of these two basic events, i.e., the method for calculating their probabilities.

[0027] To calculate the probabilities of the aforementioned basic events, we first need to determine the total failure probability of software common-cause failures. Among the multiple instrumentation and control platforms in a digital instrumentation and control system, the total failure probability U of software common-cause failures for any given platform can be calculated according to Equation 1 below.

[0028] [Formula 1]

[0029] U = ( n · P1 · V1 · V2 + n · P2 · V3 ) · P3 · P4

[0030] In Equation 1, n is the number of lines of code in the software module, P1 is the probability of an error occurring in each line of code, V1 is the probability of no code error being found during independent analysis, V2 is the probability of no code error being found when the computer tests a high-level language program, V3 is the probability of no code error being found when the computer tests a machine language program, P2 is the probability of an error occurring in each line of code due to the compiler, P3 is the probability that the code error will cause the entire system to completely crash, and P4 is the probability of a signal input that triggers a specific error code under a specific operating condition and causes the software to fail due to a common cause.

[0031] The values ​​of various parameters in Equation 1 can be obtained from the modeling input obtained in step S1. For example, n can be obtained from code measurement tools, P1 to P4 can be obtained from software history reports, and V1 to V3 can be obtained from code injection experiments.

[0032] Here, as an example, for the instrumentation and control platform of the digital instrumentation and control system in a third-generation nuclear power plant, in the calculation of the total failure probability U, n is taken as 1,000,000, P1 as 1E-3, P2 as 1E-6, P3 as 0.015, P4 as 5E-3, and V1, V2, and V3 as 0.01. This yields the total probability of software common-cause failure. For digital instrumentation and control platforms in other nuclear power plants, the values ​​shown in the example can be used as a reference, or the values ​​can be determined based on relevant statistical data using the methods described above.

[0033] Next, the probability calculation methods for basic events of common-cause software failures between instrumentation and control platforms and for basic events of common-cause software failures within instrumentation and control platforms will be explained.

[0034] Specifically, common software failures between instrumentation and control platforms refer to the simultaneous failure of multiple independent instrumentation and control platforms (such as the reactor protection system instrumentation and control platform and the power plant control system instrumentation and control platform) due to sharing the same software defects. These defects may originate from shared development toolchains, the same algorithm libraries, the same requirements specifications, or the same third-party components. Even if the platforms are completely isolated in hardware, they may still lose their independence due to commonalities at the software level.

[0035] Common-cause failure of software within an instrumentation and control platform refers to the simultaneous or successive failure of the same instrumentation and control platform due to a common software defect. Such defects usually originate from the platform's unique operating system, compiler, communication protocol stack, or application software, and do not directly affect other independent instrumentation and control platforms.

[0036] In this embodiment, for an instrumentation and control platform, the basic events of common software failures between instrumentation and control platforms and the basic events of common software failures within an instrumentation and control platform are regarded as mutually exclusive events. That is, the sum of the probabilities of the basic events of common software failures between instrumentation and control platforms and the basic events of common software failures within an instrumentation and control platform is equal to the total probability U of common software failures of the instrumentation and control platform.

[0037] Based on this, the probability of the basic event of common cause failure of software between instrumentation and control platforms, i.e., the probability U1 of common cause failure of software between instrumentation and control platforms, is calculated by the following Equation 2.

[0038] [Equation 2]

[0039] U1 = π · U

[0040] In Equation 2, π represents the proportion of the probability of common-cause software failure among instrumentation and control platforms to the total probability of common-cause software failure across all instrumentation and control platforms. The value of π can be selected based on the relationship between the technical methods and architectures of the multiple instrumentation and control platforms.

[0041] For example, if an instrumentation control platform uses a different technical approach than other instrumentation control platforms (e.g., one uses a CPU or similar technical approach, while other instrumentation control platforms use FPGA or similar technical approaches), then it can be assumed that the instrumentation control platform will hardly be affected by software defects in other platforms, nor will it be affected by software defects in other platforms. Its common software failures will only occur within the platform. Therefore, in this case, π is 0.

[0042] If an instrumentation and control platform uses the same technical methods as other instrumentation and control platforms but has a different architecture, it can be considered that the platform may share some defects with other platforms, but the main common software failure risks are still unique to itself. Therefore, in this case, π can take a relatively small value, such as 0.05. The specific value can be selected by referring to relevant statistical data.

[0043] If an instrumentation and control platform uses the same technical methods and architecture as other instrumentation and control platforms, then it can be assumed that software defects on that platform will almost certainly appear on other platforms, and vice versa. Therefore, common software failures between platforms are the main risk. Thus, in this case, π is set to 1.

[0044] Corresponding to the calculation of U1 above, the probability of the basic event of common cause failure of software within the instrumentation and control platform, i.e., the probability of common cause failure of software within the instrumentation and control platform U2, is calculated by the following Equation 3.

[0045] [Formula 3]

[0046] U2 = (1 - π) · U

[0047] Therefore, through Equations 1 to 3 above, the probability calculation of basic events of common-cause software failures between instrumentation and control platforms and basic events of common-cause software failures within instrumentation and control platforms was completed.

[0048] Next, in step S3, fault tree modeling is performed. Specifically, in step S3, basic events representing probabilistic software common-cause failures between instrumentation and control platforms and software common-cause failures within the instrumentation and control platform are added to the fault tree.

[0049] Figure 2 An example of constructing a fault tree model for an instrumentation and control system failure is shown. For example... Figure 2 As shown, when constructing the fault tree for the instrumentation and control system, "instrumentation and control system failure" is first set as the top event, and its next level is expanded using a logical OR gate. Based on the physical and logical causes of the instrumentation and control system failure, the top event includes at least two branches: instrumentation and control system software failure and instrumentation and control system hardware failure. Next, under the OR gate branch for instrumentation and control system software failure, two probabilistic basic events are added: common software failure between instrumentation and control platforms and common software failure within the instrumentation and control platform. The probability of these two basic events is specifically calculated using the method described in step S2.

[0050] It should be understood that, Figure 2 The example shown is just one instance of fault tree modeling. Fault tree modeling can be specifically modeled according to the actual analysis event object and the structure of the instrumentation and control system. For example, an instrumentation and control system failure can also be an event branched from a higher-level top event, and an instrumentation and control system software failure can also branch into a common cause failure of the instrumentation and control software, with the two basic events mentioned above added under the common cause failure of the instrumentation and control software branch.

[0051] Finally, in step S4, fault tree analysis is performed based on the fault tree model constructed in step S3.

[0052] In detail, in step S4, fault tree analysis software such as CAFTA and OpenFTA is used to analyze and assess the overall impact of common software failures on the top event (instrumentation and control system failure). This analysis process includes steps such as solving the minimum cut set for qualitative analysis, calculating the probability of the top event for quantitative analysis, and importance analysis.

[0053] The following example illustrates the evaluation method performed through steps S1 to S4 described above.

[0054] Assume that the digital instrumentation and control system of a nuclear power plant includes a reactor protection system instrumentation and control platform (hereinafter referred to as "Platform A") and a power plant control system instrumentation and control platform (hereinafter referred to as "Platform B"). Among them, Platform A and Platform B are instrumentation and control platforms with the same technical methods but different architectures.

[0055] According to step S2, the total probability U of common cause failure of software on platform A is calculated, for example, U = 1.0E-5. Taking π as 0.05, the probability U1 of common cause failure of software between instrumentation and control platforms is 5.0E-7, and the probability U2 of common cause failure of software within instrumentation and control platform is 9.5E-6.

[0056] The top event of the fault tree used for analysis is set to "Reactor protection function failure". After modeling, two basic events are set below the OR gate of "Instrumentation and Control System Software Failure": E1: "Common software cause failure between Platform A and Platform B" and E2: "Common software cause failure within Platform A". The probability of event E1 is U1, and the probability of event E2 is U2. Simultaneously, the fault tree also includes independent hardware failures and other independent software failures as basic events.

[0057] Next, the fault tree is input into the analysis software for analysis.

[0058] In qualitative analysis, the software solves for all minimal cut sets of the top event. The software outputs multiple cut sets, such as cut sets containing only a single hardware failure basic event, cut sets containing only E2 (meaning that the top event can be caused by a common cause failure of software within platform A), and cut sets containing both E1 and failure events of other platforms B.

[0059] By comparing the cut set composition, it is possible to identify whether a common cause of software failure occurs in a first-order cut set (i.e., triggers the top event alone), thereby determining its importance.

[0060] In quantitative analysis, the software calculates the probability of the top event based on the probability values ​​of each basic event using disjoint or approximate algorithms. In this example, without considering common software failures, the probability of the top event is assumed to be P_noCCF = 1.0E-4. However, after adding E1 and E2, the recalculated probability of the top event is assumed to be P_withCCF = 1.2E-4. Therefore, the incremental risk brought about by common software failures can be quantified as ΔP = 2.0E-5, a relative increase of 20%. This increment likely mainly comes from the first-order contribution of E2 (9.5E-6) and the contribution of E1 combined with related hardware events.

[0061] In the importance analysis, the analysis software was used to calculate the importance index (such as FV importance and risk amplification factor) for each basic event. For example, the calculated FV importance of E1 was 0.003, and the FV importance of E2 was 0.08. This indicates that E1 contributes approximately 0.3% to the risk of the top event, while E2 contributes 8%. The analysis results suggest that priority should be given to the software diversity and validation adequacy within platform A (reducing E2), while the common factor between platforms (E1) is acceptable given that its probability has been controlled at a low level by setting π to 0.05.

[0062] In addition to the above analysis, uncertainty analysis and sensitivity analysis can also be performed. Since the probability values ​​of E1 and E2 depend on empirical parameters such as π, the parameter variation range can be further set in the analysis software (e.g., π can be set to 0.02 to 0.1) to observe the fluctuation of the top event probability. If it is found that the top event probability is highly sensitive to the value of π, then a more in-depth evaluation of the differences in technical methods between platforms is needed, or the rationality of the value of π needs to be further demonstrated.

[0063] (Main effects of this implementation method)

[0064] Based on the evaluation method of this implementation, a technical process is established that includes four steps: modeling input acquisition, basic event value calculation, fault tree modeling, and fault tree analysis. Furthermore, the calculation of basic event values ​​uses quantitative calculation formulas rather than subjective assignment based on experience. Therefore, it is possible to evaluate software common cause failures in a systematic and standardized manner, simplifying the analysis process of software common cause failures in the PSA of nuclear power plants and improving the quality and reliability of modeling and quantitative analysis of software common cause failures in the PSA.

[0065] Furthermore, the evaluation method of this implementation provides a multi-factor comprehensive model and clear parameter value rules, offering a reliable data acquisition approach that is engineering-operable and results-reproducible, avoiding the dilemma of relying on simple subjective assignments by experts or waiting for industry-standard data.

[0066] Furthermore, the evaluation method in this embodiment utilizes the share factor π and the design of mutually exclusive basic events, ensuring that the two types of events in the quantitative analysis of the fault tree will not simultaneously appear in the same minimal cut set, thus improving the rationality and accuracy of the probability calculation results. Moreover, based on the differences in technical methods or architectures between platforms, a discrete value rule for π is provided, allowing parameter setting to be completed without complex calculations, facilitating rapid application by engineers.

[0067] The evaluation method according to one embodiment of the present invention has been described above. However, the present invention also provides an electronic device for performing the above-described evaluation method. This electronic device includes a processor, a memory, and an internal communication bus. Specifically, the processor and the memory interact via the internal communication bus; the memory stores computer instructions executable by the processor, including program code for implementing steps S1 to S4 of the aforementioned evaluation method. When the processor executes these instructions, it can sequentially complete the acquisition of modeling inputs, the calculation of basic event probabilities, the construction of a fault tree, and quantitative analysis based on the fault tree.

[0068] In addition, the electronic device may include a communication port for exchanging data with external devices (such as user terminals, PSA analysis servers, or data acquisition systems), for example, receiving modeling inputs such as instrumentation and control system manuals and test data, and outputting analysis results (such as top event probability, importance index, minimum cut set, etc.). In some embodiments, the electronic device may also include a non-volatile storage medium such as a hard disk for long-term storage of modeling parameters, fault tree structures, and historical analysis data.

[0069] Furthermore, this invention also provides a computer-readable storage medium storing computer program code thereon. When the computer program code is executed by one or more processors, it can implement all or part of the steps in the above-described evaluation method. The storage medium can be any form, such as read-only memory (ROM), random access memory (RAM), optical disc, USB flash drive, or portable hard drive, and is suitable for deployment and distribution in probabilistic safety assessment toolchains in the nuclear power field to ensure the consistency and reproducibility of the analysis process.

[0070] Furthermore, this invention also provides a computer program product comprising computer program code. When the computer program code is loaded and executed by a processor, the processor is able to perform the steps in the aforementioned evaluation method, including: acquiring modeling input, calculating the probability of basic events, establishing a fault tree containing basic events of software common-cause failures within and between platforms, and performing quantitative analysis using the fault tree. This computer program product can be provided as a software installer or a dynamic link library, facilitating integration into existing PSA software platforms (such as CAFTA, OpenFTA, etc.), thereby significantly improving the evaluation efficiency and reliability of results for software common-cause failures in digital nuclear power plants.

[0071] It should be understood that within the scope of this invention, the various parts of the embodiments can be freely combined, or the various parts of the embodiments can be appropriately modified or omitted.

Claims

1. A method for common cause failure evaluation of software of a nuclear power plant digitalized instrumentation and control system, said instrumentation and control system comprising at least two instrumentation and control platforms, characterized in that, The evaluation method includes the following steps: Obtain modeling input; Based on the modeling input, calculate the probability of the basic event used for modeling; Establish a fault tree, the fault tree including the basic events; and The failure probability of the digital instrumentation and control system is analyzed using the fault tree and the probability of the basic events. For any instrumentation and control platform, the basic events include: Basic events of common-cause software failure between instrumentation and control platforms, wherein the basic events of common-cause software failure between instrumentation and control platforms are basic events in which a common-cause software failure occurs between this instrumentation and control platform and other instrumentation and control platforms; and Basic events of common-cause software failure within the instrumentation and control platform: These are the basic events in which a common-cause software failure occurs within the instrumentation and control platform. Furthermore, the common-cause failure event between the instrumentation and control platforms and the common-cause failure event within the instrumentation and control platform are mutually exclusive events.

2. The evaluation method for common-cause failure of digital instrumentation and control system software in nuclear power plants as described in claim 1, characterized in that, Calculating the probabilities of the basic events used for modeling includes calculating the total probability of common-cause failures of the instrumentation and control platform software. For any instrumentation and control platform, the total probability of common-cause failures of the instrumentation and control platform software is equal to the sum of the probabilities of common-cause failures of the basic events between the instrumentation and control platforms and the probabilities of common-cause failures of the basic events within the instrumentation and control platform. The total probability of common-cause failure of the instrumentation and control platform software is calculated using the following formula: U = ( n · P1 · V1 · V2 + n · P2 · V3 ) · P3 · P4, In the formula, U is the total probability of common cause failure of the instrumentation and control platform software, n is the number of lines of code in the software module, P1 is the probability of an error occurring in each line of code, V1 is the probability of no code error being found during independent analysis, V2 is the probability of no code error being found when the computer tests the high-level language program, V3 is the probability of no code error being found when the computer tests the machine language program, P2 is the probability of an error occurring in each line of code due to the compiler, P3 is the probability that the code error will cause the entire system to completely collapse, and P4 is the probability of a signal input that triggers a specific error code and causes common cause failure of the software under a specific operating condition.

3. The evaluation method for common-cause failure of digital instrumentation and control system software in nuclear power plants as described in claim 2, characterized in that, Calculating the probability of the basic events used for modeling includes calculating the probability of the basic events of common-cause software failures between the instrumentation and control platforms using the following formula: U1 = π · U, In the formula, U1 represents the probability of a basic event of software common-cause failure among the instrumentation and control platforms, and π represents the proportion of the probability of a basic event of software common-cause failure among the instrumentation and control platforms to the total probability of software common-cause failure among the instrumentation and control platforms. For any instrumentation and control platform, When the technical methods of this instrumentation and control platform are different from those of other instrumentation and control platforms, π takes the value of 0. When the instrumentation and control platform uses the same technical methods as other instrumentation and control platforms but has a different architecture, the value of π is 0.

05. When the instrumentation control platform uses the same technical methods and has the same architecture as other instrumentation control platforms, π is set to 1.

4. The method of claim 1, wherein the method is used for evaluating common cause failures of software in a nuclear power plant digital I&C system. The steps involved in building a fault tree include: The failure of the instrumentation and control system is designated as the top event. The top event shall include at least two branches: instrumentation and control system software failure and instrumentation and control system hardware failure. Under the OR gate branch of the instrumentation and control system software failure, add the basic event of common cause failure of software between instrumentation and control platforms and the basic event of common cause failure of software within the instrumentation and control platform.

5. The evaluation method for common-cause failure of digital instrumentation and control system software in nuclear power plants as described in claim 2 or 3, characterized in that, The modeling input includes the test data of the instrumentation and control system. The parameters used to calculate the total probability of common-cause failure of the instrumentation and control platform software are obtained from the test data.

6. An electronic device comprising a processor and a memory for storing instructions executable by the processor, characterized in that, The processor is used to execute the instructions to implement the evaluation method for common cause failure of the digital instrumentation and control system software of a nuclear power plant as described in any one of claims 1 to 5.

7. A computer storage medium storing computer program code, characterized in that, The computer program code, when executed by a processor, implements the evaluation method for common-cause failure of the digital instrumentation and control system software of a nuclear power plant as described in any one of claims 1 to 5.

8. A computer program product, comprising computer program code, characterized in that, When the computer program code is executed by the processor, the processor implements the steps in the evaluation method for common cause failure of the digital instrumentation and control system software of a nuclear power plant as described in any one of claims 1 to 5.

Images