从数据到决策：资产管理中的人工智能I.pdf

1、Wednesday,January 8,202510:45 PM 12:15 PMUSING AI FOR REMAINING USEFUL LIFE PREDICTION IN POWER GENERATIONMEET YOUR SPEAKERSENERGY LEAD FOR HYBRID INTELLIGENCE CAPGEMINI ENGINEERINGDrew SwansonDLEAD DATA SCIENTIST CONSULTANTCAPGEMINI ENGINEERINGJeremy EvansJThe information contained in this document

2、is proprietary.Copyright 2024 Capgemini.All rights reserved.|3EDIST 2024|Using AI to Advance Predictive MaintenanceHYDROPOWER|WHY THIS MATTERS2nd Largest Renewable Source7%of Baseload 102 GW of CapacityUnited States3CanadaLargest Renewable Source60%of Baseload 82 GW of CapacityAging InfrastructureCr

3、ucial to Meeting Sustainability GoalsOPG BEARING HEALTHASSET VALUE FACTSRUL completed for OPGs R.H.Saunders Hydroelectric Generating Station.Station produces 1045 MW across 16 units(or approx.65 MW/unit)Downtime for a thrust bearing failure is a forced outage that can put a unit out of revenue produ

4、ction for 8 months*!One day lost on one unit is$100,000/day.Each unit generates approx.$35M/year in revenue.*https:/ Confidential Capgemini 2023.All rights reserved|4Some initial values from OPG website used in our internal summary,but feel free to do with this what you want.Brian can you validate?S

5、peak to driver for this particular studyThe information contained in this documentis proprietary.Copyright 2024 Capgemini.All rights reserved.|5EDIST 2024|Using AI to Advance Predictive MaintenanceTo reduce operation and maintenance costs,Hydro power plants are moving from corrective and periodic ma

6、intenance to predictive maintenance strategies.Such a transition requires changes in the data that needs to be retrieved and the type of decision processes to be employed.Advanced monitoring and data analysis technologies are essential for supporting predictive strategies.They can provide precise in