Prognostics
Framework Run-Time Software: Health Management System
Operational Software
The framework's operational
software must process the outputs of prediction mechanisms and status
sensors to provide output useful for both operation and maintenance. Software requirements include:
1. Accept both discrete and continuously valued inputs
from prediction mechanisms and sensors
2. Analyze all the inputs as specified during
development so as to provide results which take into account all the
information available
3.
Generate outputs as required for: time-based
prediction of component failure, time-based prediction of
function failures and prediction of mission success for any
profiled mission
4. Be portable for hosting on various embedded
processors
5. Require relatively small amounts of read-only
memory (ROM) and random access memory (RAM) to minimize hardware penalties
6.
Be able to operate with no
disk drive or file system
7. Have clearly defined interfaces for both input data
and output data to support all equipment specific applications
8. Consist of a generic software package and
customizing tables to simplify development and maintenance of framework
implementations.
The Prognostics Operational
Software has three interfaces as follows:
1.
Equipment Information Inputs
from the prediction mechanisms and the equipment sensors through equipment
sensor drivers
2.
Equipment Display Outputs to equipment operators
and maintainers though equipment drivers and hardware
3.
Prognostics Framework Customizing Tables
provided from the Prognostics Framework Development Software
Both
interfaces 1 and 2 must be customizable.
Interface 1 must include elements which identify which input data
is valid and which is not, and may contain equipment information in
Boolean, integer, floating point or string format as appropriate to the
prediction mechanisms and sensors. Interface
2 must be customizable to provide results in one or more of the following
formats: integer tables, lists of text strings, function return values or
failure event calls to external routines.
To implement a prognostics
prediction, the prognostics framework operational software will implement
a six-step process as shown below.
Each
of these steps is implemented by a software design element as summarized
below:
1.
Get
Prediction Mechanism Values and Sensor Values
- This
software element uses the customized tables to access the data from a
sensor or prediction mechanism and provide it in a usable format to other
elements in the prognostics framework operational software
2. Interpret
the Implication of Each Input Value on Each Future Time Point - This software element interprets the input values
into pass, fail or sometimes values for each fault and each future time
3. Diagnose
Failures for Each Future Time Point
- This
software element applies the Diagnostician technology to each future time
point to predict the faults that will occur
4. Extract
the First Failure Times For Each Predicted Failure - This software element removes the extraneous data
associated with the continuation of a fault through time points after its
first predicted occurrence.
5. Identify
The Components, Functions and Mission Profiles Affected By The Predicted
Failures - This
software module interprets the implications of predicted faults for
equipment missions, functions and maintenance
6. Provide
Operator and Maintainer Outputs
- This
software element uses the customized tables to provide the predicted
faults and fault implications to the operators and maintainers in a usable
manner.
Prognostics
Framework development tool suite
Prognostics
Framework Run-Time Software: Health Management System
Preprocessing Techniques for Prognostic Framework Inputs
Download
Prognostics Framework Health Monitoring System Demonstration
Return
to Products Page