Rotating body monitoring and alert system and method
A system and method as disclosed herein develops a predicted current remaining useful life (RUL) of a component through a generalized fault and usage model that is designed through a process of simplifying Paris' Law (or other power law) in conjunction with a Kalman Smoother (or other filtering technique). One of the many advantages of this state observer technique is that the backward/forward filtering technique employed by the Kalman Smoother has no phase delay, which allows for the development of a generalized, zero tuning model that provides an improved component health trend, and therefore a better estimate of the predicted current RUL.
1. A monitoring and alert system for estimating the remaining useful life of a rotating component of a machine, the system comprising:
a sensor positioned and configured to measure a load on the rotating component, the sensor sending a signal representative of the measured load;
a processing module configured to receive the signal, the processing module including a processor executing a set of instructions for:
developing a health indicator (HI) of the rotating component based upon the signal and a condition indicator of the rotating component, wherein the condition indicator is an assessment of nonconformity of the rotating component; and
developing a prediction of remaining useful life (predicted RUL) of the rotating component based upon the HI;
determining a dHI/dt, a dRUL/dt, and a d 2 RUL/dt 2 ;
applying a first Kalman filter to the HI and the dHI/dt determinations, wherein the first Kalman filter is an anti-causal backward/forward filter;
applying a second Kalman filter to the predicted RUL, the dRUL/dt, and the d 2 RUL/dt 2 determinations; and
determining a measure of confidence in the predicted RUL based on the dRUL/dt and the d 2 RUL/dt 2 determinations.
2. The monitoring and alert system according to claim 1 , wherein the HI is developed using a whitening linear transform.
3. The monitoring and alert system according to claim 1 , wherein the sensor includes one or more of a strain gauge, a vibration sensor, and an acoustic emission sensor.
4. A method of predicting the remaining useful life of a rotating component of a vehicle comprising:
receiving data representative of a load on the rotating component and a condition of the rotating component from a sensor on the vehicle, wherein the condition is an assessment of nonconformity of the rotating component;
generating a health indicator (HI) based upon the received data;
predicting a remaining useful life (predicted RUL) of the rotating component based upon the HI;
evaluating the HI and RUL so as to determine whether to alert an operator as to the condition of the rotating component, wherein the evaluation includes:
determining a dHI/dt, a dRUL/dt, and a d 2 RUL/dt 2 ;
applying a Kalman Smoother to the HI and the dHI/dt determinations;
applying a Kalman Filter to the predicted RUL, the dRUL/dt, and the d 2 RUL/dt 2 determinations; and
determining a measure of confidence in the predicted RUL based on the dRUL/dt, and the d 2 RUL/dt 2 determinations.
5. The method according to claim 4 , wherein the generating is performed by using a whitening linear transform.
6. The method according to claim 4 , wherein the dRUL/dt is compared against a value of −1 and the d 2 RUL/dt 2 is compared against a value of 0.
7. The method according to claim 6 , wherein, when the HI is above about 0.5, the dRUL/dt is approximately −1, and the d 2 RUL/dt 2 is approximately 0, the operator is alerted as to the condition of the rotating component.
8. The method according to claim 7 , wherein the predicting uses a surrogate for range of strain that is a measure of predicted load on the rotating component.