建築熱源システムを対象としたデータ駆動型不具合検知・診断　（第3報）熱源システムシミュレーションを用いた不具合の教師データベースの作成と分析

抄録

We developed a simulation to calculate the system behavior affected various faults and generated training database of faults using this simulation to examine fault detection and diagnosis (FDD) by convolutional neural network (CNN). And we analyzed the characteristics of the effects of faults and examined a method to extend the database efficiently.