|
The problem for facility planning is considered for get minimize
cost of the facility layout. There have been many
other proposed applications of these, forexample, factory layout
problems, circuit layout on PCB, plannar facility layout,
information retrieval for storage facility and site position in
network., etc. We can find this kind problem can be modelled as
a Quadratic Assignment Problem (QAP).Quadratic assignment
problem is a discrete, nonlinear, nonconvex problem. Sahni and
Gonzale have been proved it belong to the class of NP-hard
problems, which are not likely to be solved exactly in the
reasonable time. And most traditional heuristic can only offer
the technical for local search method. Get the cost can only be
optimal solution.In this research, the main purpose is try to
solve the problem of facility planning, using these AI heuristic
that developed in recently, Threshold Accepting (TA), Record to
Record Traveling (RRT), to be the kind technology for global
searching. Use testing problems in QAPLIB for test AI heuristic
efficient and cost. When considering the practical problem will
meet the situation for site constraint , therefore we make a new
mathematical model for solve this problem. After modelling, use
an example from factory of the mass rapid transit, then use the
AI heuristics that we developed for solve this example. lity
planning is considered for get minimize cost of the
|