多目标的模式识别优化法及其在V－PTC材料设计中的应用

Pattern Recognition Optimization(PRO)Method for Multiple Targets and Its Application to V-PTC Materials Design

利用多目标的模式识别优化法进行计算机辅助Ｖ－ＰＴＣ材料设计，通过主成分分析研究了Ｖ－ＰＴＣ材料的性能与其特征参数的关系，得到了以Ｖ－ＰＴＣ材料的ρ０／ρｎ、ρｍ／ρｎ、和ρｒ为目标变量的模式识别分类图，并用来探寻具有较好性能的样品的工艺条件和配方参数。根据多目标的模式识别优化法提供的信息，设计并合成了数个优化样品。

The computer-aid design of V-PTC materials was investigated by pattern recognition optimization(PRO)method for multiple targets.The relationships between the properties and their features of V-PTC materials were studied through the principal component analysis.The classification diagrams with the targets including ρr,ρ0/ρn and ρm/ρn were obtained and used to explore the technical conditions and compositions of the samples with better properties.Based on the information provided by PRO method,a few optimal samples of V-PTC were synthesized in our laborotary.