基于改进粒子群算法的混凝土坝热学参数反演研究

Inverse analysis of concrete dam thermal parameters based on an improved particle swarm optimization method

为实时获取真实浇筑环境下混凝土热学特性变化规律,基于温度观测数据,建立了材料参数识别的群体智能优化粒子群模型。针对传统粒子群算法易早熟陷入局部极值点这一问题,将粒子群算法与凹函数权值递减策略相结合,对粒子群算法进行改进,通过算法性能对比分析来说明其有效性。考虑冷却通水以及外界气温变化影响,将改进的粒子群算法用于某拱坝高温季节热学参数反演分析,通过实测温度与计算温度对比分析来说明反演参数的合理性。算例分析表明,改进的粒子群算法具有收敛速度快、识别精度高等优点。

A swarm intelligence optimization algorithm-particle swarm optimization(PSO)model was constructed for thermal parameter identification based on temperature observation data.Therefore,thermal characteristics of concrete in real pouring environment can be obtained in real time.In order to solve this problem that traditional PSO is easy to fall into the local extreme point prematurely,an improved particle swarm optimization(IPSO)was proposed,which combines particle swarm optimization and concave function weight decreasing strategy.Effectiveness of IPSO was illustrated by the algorithm performance comparison analysis.Considering the influence of cooling water and outside temperature change,IPSO was applied to inverse analysis of thermal parameters of an arch dam during hot season construction.The inversion parameters were validated by comparison of measured temperature and calculated temperature.The numerical example shows that IPSO has advantages of converging quickly and identifying accurately.