Helicopter performance for the U.S. Navy
The U.S. Navy had prepared extensive tables that allow computation
of the lift produced by a helicopter rotor. There are as many as
five variables that need to be taken into account:
- Rotor speed
- Rotor pitch
- Rotor tilt
- Air temperature
- Air pressure
The Navy wanted to develop a software program to facilitate this
calculation, but implementing a table lookup produced a large and
complex program, plus it did not allow for analysis of the relationships
among these variables.
The solution was to use MPR to generate a single equation that
correlated the effect of all of these variables on lift. The application
of MPR to this problem was described as "very successful."
Modeling and control of advanced life support systems
for travel to Mars - NASA
NASA funded development of models to help with the control of advanced
life support systems for possible future use on Mars. The models
were used to predict gas exchange from crops in simulated Mars colony
greenhouses.
|
Water Filtration Plant Performance
The New Jersey Department of Environmental Protection funded a
project to develop models to predict the performance of water treatment
filters. The models were used to predict turbidity that escapes
the filter, based on measurements of particle size distribution
entering the filter.
Demand forecasting for a natural gas provider in Europe
The energy company Blugas, SpA. in Mantua, Italy used MPR to develop
models for natural gas demand forecasting. They needed to estimate
the amount of gas they would need to provide to their customers
the next day so they could optimize the amounts they would have
to buy and store. The predictors included the forecast temperature,
the gas demand on the current day, and day of the week.
|