Bollinger fuel algorithm

Takeaway: The Bollinger fuel algorithm is one strategy that
answer the question of refueling a car (a truck, or any
vehicle that has a known capacity for the fuel and
some user-provided information about how much the
vehicle consumes). The Bollinger comes from a well-known financial technical
analysis instrument called Bollinger’s bands. In this article we will be investigating
the use of Bollinger’s bands in refueling strategies.

Problem statement:

“Given uncertain future prices of a commodity, when
should one replenish a fuel tank of a car?”
Limited by: one has to use the car, ie. the tank cannot be a negative number, and
in practise it cannot be even close to zero (too risky).

Bollinger’s bands to the rescue!

Bollinger’s bands are a method to estimate the likelihood
of a current minimum or maximum, based on the variance of
a commodity (a price) in the last D samples.

In this application we try to find out whether using
these limits are useful in deciding when and how much to
take fuel into a car’s fuel tank.

=> local minima exploited
=> local maxima avoided

- does fuel as commodity have the same rules as a stock price (or index)?
- economic shocks
- how much does the car's fuel capacity affect decisions? And in what 
  way?

Algorithm [prototype draft]

function BREP(F,stn[],C)

 - Bollinger replenishment algorithm
 - given Fuel (=amount)
 - stn[] list is a list of [x,y,price] stating the known 
   prices in an area
 - C is a consumption function of some kind
   -> sets limit to the practical search cutoff value
      (can't search forever if there's a real consumption)