HUMS Optimal Weighting of Condition Indicators to Determine the Health of a Component

Integrated Mechanical Diagnostics Health and Usage Management Systems (IMD-HUMS) give the maintainer information regarding the mechanical condition of a component. The new Sikorsky S-92 medium lift helicopter incorporates the Goodrich IMD-HUMS, which monitors the health of the drive train components, such as the main transmission, and intermediate and tail rotor gearboxes. This system computes a variety of condition indicators (CI) from physical measurements, such as vibration. A number of CIs are fused into the Health Indicator (HI), which is scaled between 0 and 1. The HI typically consists of four to six condition indicators. Not all CIs are sensitive to every component failure mode. By fusing a number of CIs, the HI can be sensitive to a number of failure modes and still have a controlled false alarm rate. Additionally, by having one HI to indicate the state of the component, the maintainer is not required to examine a large number of disparate CIs or have knowledge of the nominal CI values. Furthermore, with the HI normalized between 0 and 1, with 0 representing a healthy component, the HI provides an easy to interpret relative indication of component health. The HI is calculated by optimally weighting the contribution of each CI, specifically by considering the distribution of all contributing CIs and then performing a transformation that orthogonalizes the CI distributions. The othogonalization process maximizes the separation between the CIs, which has the effect of maximizing the probability of event detection for a given probability of false warning. Optimality is proven by showing that the derivative of the transformation matrix is zero. Further, it is shown that the transformation matrix is the product of the Eigenvector matrix with the covariance Eigenvalues.