SIMULATED ANNEALING

In static Geologic Modeling and Dynamic Simulation the reservoir heterogeneity
description represents one of the most critical phases.
How much a modeling method can approach the reality can be only statistically
defined.
If we consider a Sequential Gaussian Simulation approach, for each input data set
an output set of multiple realizations will be calculated.
In this context further results add new constraints to define a statistical solution.
Simulated Annealing like the Sequential Simulation method honors the univariate
statistics, spatial relationship and relationships among different attributes.
It also has the flexibility to incorporate other contraints.
The simulation Algorithm that incorporates the constraints is described by the
Objective Function.
Alternatively we can dynamically perform such a kind of simulation through
Neural Networks with the Hopfield Net.
Through the competitive logic of this network, optimal results can be achieved.
Simulated Annealing can be optimally used to describe heterogeneity in vulcanic
formations or carbonate diagenetical evolution.