混凝土浇筑仓温度双控指标拟定的最大熵法

结合西南某建设中的混凝土特高拱坝高温季节浇筑仓实测温度,采用最大熵法拟定了混凝土浇筑仓温度双控指标,即混凝土浇筑仓达到最高温度前典型龄期对应的容许温度和容许温度变化率。为了获得高精度的最大熵概率密度函数,采用粒子群算法来优化求解拉格朗日乘子系数。工程实例分析表明:最大熵法不需要事先假设样本分布类型,直接根据各基本随机变量的数字特征值进行计算,就可以得到精度较高的概率分布密度函数,由此求出的温度双控指标是可行的。

Maximum Entropy Method of Drafting Double-Index of Temperature Control for Concrete Pouring Surface

The double-index of temperature control for concrete pouring surface, i.e. the corresponding allowable temperature and temperature gradient at typical age before the pouring surface reaches its maximum temperature, was obtained by maximum entropy method (MEM). The measured temperature of the concrete pouring surface of an extra-high arch dam in high temperature season in southwest China was taken as background in this research. In order to obtain highly accurate maximum entropy probability density function, particle swarm optimization (PSO) was adopted to optimize and solve the Lagrange multiplier coefficients. It is indicated by the engineering practice analysis that sample distribution type needs not to be assumed in advance for MEM, and highly accurate probability distribution density function can be obtained by the calculation of the numerical eigenvalue of each basic random variable. The obtained double-index is feasible.