System and method for detecting lubricated bearing condition

A method of operating an analytical engine system, comprising:

• receiving a lubricant temperature via at least one of a plurality of sensors disposed on an engine system;
• loading model data from a memory, wherein the model data is based at least in part on sensor feedback during prior operation of the engine system;
• determining a model probability distribution based on the model data, wherein the model data comprises model lubricant temperature data and model load data;
• determining a distance threshold value of the model probability distribution based at least in part on a threshold percentage;
• determining a window probability distribution based on window data sampled during operation of the engine system, wherein the window data comprises window temperature data having the lubricant temperature and window load data that is based at least in part on feedback from the plurality of sensors during current operation of the engine system;
• determining a fraction of the window probability distribution that is greater than the distance threshold value;
• generating a lubricant alert signal when the fraction is greater than a temperature anomaly threshold of between 0 and 1, wherein the temperature anomaly threshold is obtained based on a desired acceptable false alarm rate, wherein the analytical engine system is configured to update the model data based on the window data when the fraction is less than the temperature anomaly threshold;
• and
• updating the model data based on the window data when the fraction is less than the temperature anomaly threshold, wherein updating the model data comprises:
• determining an updated model probability distribution based on the model data and the window data;
• determining an updated distance threshold value of the updated model probability distribution based at least in part on the threshold percentage;
• and
• replacing the model probability distribution with the updated model probability distribution and replacing the distance threshold value with the updated distance threshold value.

A non-transitory computer readable medium comprising instructions configured to be executed by a processor of a control system, wherein the instructions comprise instructions configured to cause the processor to:

• receive a lubricant temperature via at least one of a plurality of sensors disposed on an engine system
• load model data from a memory, wherein the model data is based at least in part on sensor feedback during prior operation of the engine system;
• determine a model probability distribution based on the model data, wherein the model data comprises model lubricant temperature data and model load data;
• determine a distance threshold value of the model probability distribution based at least in part on a threshold percentage, wherein the threshold percentage is greater than or equal to 90 percent;
• determine a window probability distribution based on window data sampled from the engine system, wherein the window data comprises window temperature data having the lubricant temperature and window load data that is based at least in part on sensor feedback during operation of the engine system;
• determine a fraction of the window probability distribution that is greater than the distance threshold value;
• generate a lubricant alert signal when the fraction is greater than a 0.5;
• and
• remove data sets of the window data to form filtered window data, wherein the removed data sets do not correspond to a steady state operation of the engine system, wherein steady state operation of the engine system is based at least in part on the window temperature data or the window load data.