An Observer-Based Solution for Structural Health Monitoring of Mechanical Systems Subjected to Unknown Inputs

The mechanical structure forms a crucial element of any engineered system as it determines the operational life of the system. For instance, the service life of an aircraft is decided by the properties of the airframe and the wings, and continuous loads acting on these structures can cause deterioration in their structural integrity. It thus becomes crucial to monitor the “health” of the structure to ensure that phenomena such as cracks can be quickly detected, and suitable corrective actions can be taken. This process is termed structural health monitoring (SHM), and it has received significant attention by industry. The SHM system involves the observation of a system over time using periodically sampled dynamic response measurements from an array of sensors and analyzing these measurements to determine the current health of the system. Traditionally, SHM has been conducted using techniques which require the system to be idle. In addition, the SHM of larger structures will require a proportionally large number of sensors to be used at certain locations in the structure. Thus, there is a need for an in-situ SHM solution which uses fewer sensors. The in-situ constraint thus requires that the structure health be monitored even in the presence of unknown inputs. To solve these problems, in this work, an unknown input observer (UIO)-based SHM solution is provided. This approach uses results from control theory and is demonstrated for a cantilever beam structure approximated as a set of two coupled spring-mass systems. It is assumed that the displacement and velocity of only one of these masses, on which the unknown input acts, are measured, and the states of the other mass are estimated based on these measurements. The advantages and limitations of the UIO-based in-situ solution are highlighted.