摘要
针对抗震结构所处的不同状态,采用有限时期马氏决策理论,探讨了设计基准期内抗震结构维修策略优化问题,建立了设计基准期内使结构期望报酬最大的随机动态规划数学模型,通过对马氏决策逆序定值的求解可以使决策者清楚明了在设计基准期内,每一检测维修阶段的最佳维修策略和总的最大期望报酬,从而为维修决策提供科学依据.对于不同类型的抗震结构,可以编制相应程序进行随机动态规划方法的迭代计算,从而简化计算过程,减少计算工作量.
According to the state of aeismatic structure, by introduction of finite period Markov decision theory, the problem of maintenance strategy optimization of aeismatie structure during design reference is discussed in detail. A mathematical model of stochastic dynamical programming with maximum expected reward during the design reference period is presented in this paper. By solution of Markov decision inverted sequence, the optimum maintenance strategy of each stage and total maximum expected reward can be obtained during the design reference period. An optimization maintenance strategy of expected reward maximization is given during the design reference period. All of these provide a scientific method for maintenance decision-making, herative computation of this stochastic dynamic method can be programmed, greatly decreasing the amount of work.
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
2006年第2期250-252,320,共4页
Journal of Harbin Institute of Technology
基金
国家自然科学基金资助项目(50378007)
北京市自然科学基金重点资助项目(8031001)
北京工业大学青年科研基金资助项目(JQ0401200372)
关键词
抗震结构
设计基准期
维修策略
马氏决策
aeismatie structure
design reference period
maintenance strategy
Markov decision