Prognostic Control and Predictive Maintenance for Variable-speed Electro-Hydraulic Drive Systems

Project Details

Description

Abstract:

In electro-hydraulic systems maintenance is a major expense as this requires a technician to state a diagnosis, eventual new components, and causes outage time which is especially costly. Furthermore, the cost increases drastically when the systems are remote such as offshore wind turbines. Thus an increasing interest in more durable and reliable solutions has sprouted. Especially, the ability to predict the need of maintenance based on the current health of the system has a potential to reduce the outage time. Unfortunately, a health sensor is not a reality, thus alternative methods are needed in order to predict maintenance. Furthermore, methods which can virtually diagnose which component is faulty in the system, could reduce the expenses regarding the technician and may prevent healthy components being replaced. In addition to this a diagnosis could be used to reconfigure the operation of the system such that the remaining useful life can be extended.
The main objective of this PhD is to be able to obtain a measure of health and based on an estimate of the remaining useful life of the system. This measure originates from an understanding of how the physics change when the systems is degrading and the ability to detect this. This will be regarded for both single- and multi axis variable speed electro-hydraulic drive systems. The expected product of this project is an algorithm that can diagnose multiple electro-hydraulic systems and based on this be able to predict maintenance, and utilize a prognosis to reconfigure the operation of the systems. Thus the outcome will be a more flexible and cost efficient maintenance strategy in the electro-hydraulic sector.


Funding: EUDP
StatusActive
Effective start/end date01/09/202131/08/2024

Collaborative partners

  • KVM
  • Bosch Rexroth
  • FLSmidth

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.