Title
Efficient pruning technique based on linear relaxations
Abstract
This paper extends the Quad-filtering algorithm for handling general nonlinear systems. This extended algorithm is based on the RLT (Reformulation-Linearization Technique) schema. In the reformulation phase, tight convex and concave approximations of nonlinear terms are generated, that's to say for bilinear terms, product of variables, power and univariate terms. New variables are introduced to linearize the initial constraint system. A linear programming solver is called to prune the domains. A combination of this filtering technique with Box-consistency filtering algorithm has been investigated. Experimental results on difficult problems show that a solver based on this combination outperforms classical CSP solvers.
Year
DOI
Venue
2003
10.1007/11425076_1
COCOS
Keywords
Field
DocType
general nonlinear system,linear relaxation,efficient pruning technique,difficult problem,quad-filtering algorithm,reformulation-linearization technique,classical csp solvers,nonlinear term,concave approximation,linear programming solver,bilinear term,extended algorithm,nonlinear system,linear program
Agronomy,Mathematical optimization,Nonlinear system,Filter (signal processing),Convex hull,Algorithm,Product term,Linear programming,Solver,Univariate,Mathematics,Bilinear interpolation
Conference
Volume
ISSN
ISBN
3478
0302-9743
3-540-26003-X
Citations 
PageRank 
References 
2
0.40
18
Authors
3
Name
Order
Citations
PageRank
Yahia Lebbah111519.34
Claude Michel210710.57
Michel Rueher361359.81