In the digital era of increasing software complexity, improving the development efficiency of safety-critical software is a challenging task faced by academia and industry in domains such as nuclear energy, aviation, the automotive industry, and rail transportation. Recently, people have been excited about using pre-trained large language models (LLMs) such as ChatGPT and GPT-4 to generate code. Professionals in the safety-critical software field are intrigued by the code generation capabilities of LLMs. However, there is currently a lack of systematic case studies in this area. Aiming at the need for automated code generation in safety-critical domains such as nuclear energy and the automotive industry, this paper conducts a case study on generating safety-critical software code using GPT-4 as the tool. Practical engineering cases from the industrial domain are employed. We explore different approaches, including code generation based on overall requirements, specific requirements, and augmented prompts. We propose a novel prompt engineering method called Prompt-FDC that integrates basic functional requirements, domain feature generalization, and domain constraints. This method improves code completeness from achieving 30% functions to 100% functions, increases the code comment rate to 26.3%, and yields better results in terms of code compliance, readability, and maintainability. The code generation approach based on LLMs also introduces a new software development process and V-model lifecycle for safety-critical software. Through systematic case studies, we demonstrate that, with appropriate prompt methods, LLMs can auto-generate safety-critical software code that meets practical engineering application requirements. It is foreseeable that LLMs can be applied to various engineering domains to improve software safety and development efficiency.

The aim of this study is to verify if the reliability of a digital four-channel RPS under the design phase satisfies the specified target and to identify the weakness of system design and potential solutions for system reliability improvement. The event-tree/fault-tree (ET/FT), which is the method used in the current probabilistic safety assessment (PSA) framework of nuclear power plants (NPPs), was adopted to developed reliability modeling for the RPS with the Top Events defined as the system failure to generate reactor trip signal and the system generating spurious trip signal. The evaluation results indicate that the probability of the system failure on demand and the frequency of spurious trip signal generation are 1.47×10⁻⁶ with a 95 upper bound of 4.63×10⁻⁶ and 7.94×10⁻⁴/year with a 95 upper bound of 2.50×10⁻³/year, respectively. The importance and sensitivity analyses were conducted and it was found that undetected unsafe common cause failures (CCFs) of signal conditioning modules (SCMs) dominate the system reliability. Two preliminary optimization schemes relative to reducing periodic test interval and adapting two kinds of diverse SCMs were proposed. Results of the quantitive evaluation of the schemes show that neither of them could determinedly improve the system reliability to the target level. In the future, more detailed optimization analysis shall be required to determine a feasible system design optimization scheme.

The safety video display unit (SVDU), as the display machine of the reactor protection system, performs the functions of displaying the reactor’s safety parameters and sending safety control commands. In order to meet the needs of nuclear power safety-level digital control system (DCS), like designing verification, operator training, and accidental drills, for the SVDU in the NASPIC platform developed independently by China National Nuclear Corporation, a virtual embedded system technology based on the micro x86 industrial host is proposed to make software simulation to the SVDU physical controller with exactly the same hardware appearance of the original one. The SVDU stimulator realized by this research can achieve 100% simulation of the logical functions of the SVDU physical equipment and the synchronized upgrading function between stimulator and the real equipment. With the development of multiple engineering application requirements, such as configuration verification and operator training, this stimulator has been applied in several virtual security level DCS projects.

Offshore floating nuclear power plant is characterized by its small and movable, which is not only able to provide safe and efficient electric energy to remote islands, but to the oil and gas platforms. The safety digital control system (DCS) cabinet, as a carrier for the electronic devices, plays a significant role in ensuring the normal operation of the nuclear power plant. To