粒子群算法对织构分析改进最大熵法的改进

IMPROVEMENT ON THE MODIFIED MAXIMUM ENTROPY METHOD FOR TEXTURE ANALYSIS BY PARTICLE SWARM OPTIMIZATION

将粒子群算法(PSO)成功地应用于改进最大熵法(MMEM),使用两张或单张部分极图数据,即可测算出多组具有宏观对称的冷轧纯铜片样品的完整晶粒取向分布(ODF).这一改进保留了原MMEM的特长,用较少的实验数据即可计算出样品的真ODF,而又消除了原法在求解非线性方程组时需要选取特定初解的困难,从而使MMEM更具普适性和实用性.

The particle swarm optimization （PSO） was successfully applied for the modified maximum entropy method （MMEM） of the quantitative texture analysis. Using two incomplete pole figures or only one incomplete pole figure, the complete orientation distributions of pure copper samples can been obtained. Such an improvement has been verified to keep the merits of primary MMEM, i.e., the complete orientation distribution function （ODF） of samples can be calculated using lesser experimental data, furthermore, the difficulty to choose a set of purposely initial solutions has also been overcome when solving the nonlinear equation, which makes the MMEM more versatile and practicable.