摘要详情

ID / 提交时间

82 / 2024-08-02 09:19:26

标题

RESEARCH ON RELIABILITY-BASED DESIGN OPTIMIZATION METHOD OF NUCLEAR PASSIVE RESIDUAL HEAT REMOVAL SYSTEM

关键字

passive residual heat removal (PRHR), reliability-based design optimization (RBDO), sequence optimization and reliability assessment (SORA), TAGA-BP, lightweigh

主题及专题

TRACK 4: MODELING & SIMULATION

状态

全文录用

作者

Yuepeng Bi / Harbin Engineering University

Genglei Xia / Harbin Engineering University

Chenyang Wang / Harbin Engineering University

Minjun Peng / Harbin Engineering University

Chang Wang / China Ship Development and Design Center

摘要

Reducing the weight of marine nuclear power plants is essential for optimal performance. Passive residual heat removal systems (PRHRS) enhance safety and are a significant trend in marine nuclear plant design. However, these systems rely on physical laws, providing less driving force than active equipment and are susceptible to uncertainties, impacting their functionality. Considering these uncertainties is crucial in design and operation. Traditional deterministic design uses safety factors to ensure reliability but lacks a solid theoretical base. To address this issue and account for uncertainty, this study focuses on reliability-based design optimization methods for passive residual heat removal systems using sequence optimization and reliability assessment (SORA). The design optimization and reliability analysis are carried out in sequence. Design optimization and reliability analysis are performed sequentially. In the design optimization part, the optimal solution of the objective function is obtained through optimization algorithm. At the optimal solution, inverse reliability analysis is carried out to solve the shifting vector through inverse reliability analysis, and the coupling of the two parts is realized. Additionally, we utilize the TAGA-BP( two-layer adaptive genetic algorithm to optimize BP neural network) model instead of the T-H code to reduce computational burden. When applied to the IP200 passive residual heat removal system, our method reduces its weight by 2.471×10³kg (3.46% reduction) while maintaining a limited probability of failure. Compared to conventional deterministic optimization approaches, reliability-based approach further optimizes safety margins. This paper's proposed method has demonstrated capability in accounting for uncertainty factors during the design optimization process and offers advantages such as fast computation speed. Overall, this approach effectively reduces weight in passive residual heat removal systems and plays a vital role in improving their overall economy.