Online Fault Identification of Digital Hydraulic Valves Using a Combined Model-Based and Data-Driven Approach
For hydraulic working machines and field robots, this approach improves fault tolerance by enabling online fault identification during normal operation.
The paper presents a combined model-based and data-driven approach for identifying faulty valves in digital hydraulic systems using pressure measurements and a flow balance model, enabling fault-tolerant operation without requiring service personnel access.
Robustness and fault-tolerance are desirable properties for hydraulic working machines and field robots. In applications where service personnel do not have easy access to the machine, it is important that the machine can continue its operation despite single machinery faults. Digital hydraulics enables a design which is inherently fault-tolerant by having each hydraulic actuator controlled by a number of parallel on/off valves. The exact state operation of a digital hydraulic system enables model based diagnostics of the hydraulic components and the possibility to compensate for identified faults. This paper presents an approach for identifying faulty valves based on combination of pressure measurements made during the normal operation of the machine and a mathematical model describing the flow balance of the hydraulic system.