A Declarative Framework for Constrained Search Problems in Manufacturing [chapter]

Sitek Pawek, Wikarek Jaroslaw
2008 Automation and Robotics  
Constrained search problems Constrained search problems (e.g., scheduling, planning, resource allocation, placement, routing) appear frequently at different levels of decision in manufacturing. They are usually characterized by technical, environmental or manpower constraints, which make them unstructured, and in most of the cases are difficult to solve (NP-complete). Traditional mathematical programming approaches (linear programming, integer and mixed integer programming) are deficient in the
more » ... following ways: their representation of constraints is artificial (commonly using 0-1 variables), their computing time in the presence of many constraints is very long (due to combinatorial explosion), and they cannot process various constraints applied to the main problem. Thus, the most used approach consists in www.intechopen.com Automation and Robotics 244 developing specific software, written in a procedural language like PASCAL, BASIC or C, to solve each particular problem. However, the use of procedural languages brings the following well known disadvantages: the development time of the programs is very long and the programs are very complex, hence difficult to maintain and adapt to rapid changes of requirements. Unlike traditional approaches, CLP provides for a natural representation of heterogeneous constraints and allows domain-specific heuristics to be used on top of generic solving techniques.
doi:10.5772/6107 fatcat:h2skdpbe2vbpnpsjrhnas6ghgq