APPARATUS FAILURE DIAGNOSIS SUPPORT SYSTEM AND APPARATUS FAILURE DIAGNOSIS SUPPORT METHOD

• Main Authors: KUDO KEIKO, NOGUCHI JUNJI
• Format: Patent
• Language: English; Japanese
• Published: 23.04.2020
• Subjects: CONTROL OR REGULATING SYSTEMS IN GENERAL; CONTROLLING; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS ORELEMENTS; PHYSICS; REGULATING

To efficiently investigate a defective part by reducing the time required to investigate the defective part in the event that a defective part investigation is performed on a facility or apparatus on the basis of sensor data.SOLUTION: A learning diagnosis system creates a diagnosis model through learning based on failure data, and stores an apparatus type, a failure cause region, and sensor data in a rare case data table in the event that the number of cases involving the failure cause region of the apparatus falls below a certain number of cases. Thereafter, an estimated probability at which each region of the apparatus having failed may become a failure cause is calculated based on the diagnosis model created by a learning unit. A sensor data coincidence ratio of the sensor data of the apparatus having failed to previous sensor data of the apparatus type is calculated based on the rare case data table. The diagnosed failure cause region of the apparatus having failed, and the estimated probability are displayed. The calculated sensor data coincidence ratio for each region of the apparatus having failed is displayed.SELECTED DRAWING: Figure 1 【課題】センサデータに基づいて、設備・機器の故障箇所調査をおこなう際に、故障箇所の究明に要する時間を短縮して、効率的に故障箇所の究明をおこなう。【解決手段】学習診断装置は、故障データから学習して診断モデルを作成し、その機器の故障原因部位として、一定の件数に満たない場合に、レアケースデータテーブルに、その機器の機種と故障原因部位とセンサデータを格納する。次に、学習部により作成された診断モデルに基づいて、故障の発生した機器の部位毎に、故障原因となる推定確率を算出し、レアケースデータテーブルに基づいて、故障の発生した機器のセンサデータと、その機器の機種の過去のセンサデータのセンサデータ一致率を算出する。そして、診断された故障の発生した機器の故障原因部位と、その推定確率を表示し、算出された故障の発生した機器の部位毎のセンサデータ一致率を表示する。【選択図】 図１