In this paper, we present an algorithm for finding utilitarian optimal solutions to Simple and Disjunctive Temporal Problems with Preferences (STPPs and DTPPs) based on Benders9; decomposition and adopting SAT tech...
详细信息
this article presents a database of about 200 graph invariants for deriving systematically necessary conditions from the graph properties based representation of global constraints. this scheme is based on invariants ...
详细信息
Scheduling is one of the most successful application areas of constraintprogramming mainly thanks to special global constraints designed to model resource restrictions. Among these global constraints, edge-finding an...
详细信息
Scheduling is one of the most successful application areas of constraintprogramming mainly thanks to special global constraints designed to model resource restrictions. Among these global constraints, edge-finding and not-first/not-last are the most popular filtering algorithms for unary resources. In this paper we introduce new O(n log n) versions of these two filtering algorithms and one more O(n log n) filtering algorithm called detectable precedences. these algorithms use a special data structures theta-tree and theta-Alpha-tree. these data structures are especially designed for "what-if" reasoning about a set of activities so we also propose to use them for handling so called optional activities, i.e. activities which may or may not appear on the resource. In particular, we propose new O(n log n) variants of filtering algorithms which are able to handle optional activities: overload checking, detectable precedences and not-first/not-last.
Interactive tasks such as online configuration and e-commerce can be modelled as constraint satisfaction problems (CSPs). these can be solved interactively by a user assigning values to variables. the user may require...
详细信息
We introduce the study of Conditional symmetry breaking in constraintprogramming. this arises in a sub-problem of a constraint satisfaction problem, where the sub-problem satisfies some condition under which addition...
详细信息
Still-life is a challenging problem for CP techniques. We show how to use the global case constraint to construct ad-hoc constraints which can provide stronger propagation than existing CP models. We also demonstrate ...
详细信息
the car-sequencing problem arises from the manufacture of cars on an assembly line. A number of cars are to be made on a production line;they are not identical because different options are available as variants on th...
详细信息
ISBN:
(纸本)3540292381
the car-sequencing problem arises from the manufacture of cars on an assembly line. A number of cars are to be made on a production line;they are not identical because different options are available as variants on the basic model. the different stations which install the various options have been designed to handle at most a certain percentage of the cars passing along the assembly line. Consequently, the cars must be arranged in a sequence so that these capacities are not exceeded. In this paper, the formulation of the car-sequencing problem is presented as a non-binary constraint satisfaction problem (CSP) withconstraints of fixed arity 5. A search algorithm based on non-binary forward checking (nPC) is used to solve the problem. For the car-sequencing problem the variables should be assigned consecutively. the choice of value ordering heuristics having a dramatic effect on solution time for this problem, different value ordering heuristics were implemented. Since any possible solution is a permutation of a fixed set of values, a succeed-first strategy for value ordering only postpones the assignment of the difficult classes and a value ordering based on fail-first could be a better choice. these methods are compared on the instances reported in the CSPLib. the results obtained showed the superiority of a strategy of fail-first type against to a succeed-first strategy. In particular, the MaxUtil and MaxPQ heuristics allowed a better exploration of the space of solutions and solved all the instances of problems with 200 variables. It should be underlined the fact that these problems were solved in little time (6 seconds on average) and the longest time is 13 seconds for the instance 90.09, whereas for ILOG Solver the least powerful time exceeds 1 minute. this result can be justified by our encoding. Indeed, we encoded the maximum of constraints (the capacity of each option, the request for each class) inside an explicit 5-ary constraint with very high tightness (
Weighted constraint satisfaction problems (WCSP) and Max-SAT are optimization versions of the CSP framework and SAT repectively. they have many practical applications. Most current state-of-the-art complete solvers fo...
详细信息
ISBN:
(纸本)3540292381
Weighted constraint satisfaction problems (WCSP) and Max-SAT are optimization versions of the CSP framework and SAT repectively. they have many practical applications. Most current state-of-the-art complete solvers for WCSP and Max-SAT problems can be described as a basic depth-first branch and bound search that computes a lower bound during the search that can be used together withthe cost of the best solution found in order to prune entire search subtrees. Recently, a collection of local consistency properties such as NC*, AC*, DAC*, FDAC* and EDAC* have been proposed for WCSP in order to simplify the problem. In Max-SAT we have recently proposed inference rules to detect unfeasible assignments. Resolution in Max-SAT is an extension of classical resolution for the SAT problem: (x ∨ A, u), (x̄ ∨ B, w) ⇒ {(A ∨ B, m) (x ∨ A, u Θ m) (x̄ ∨ B, w Θ m), (x ∨ A ∨ B̄, m) (x̄ ∨ Ā ∨ B, m) where A and B are arbitrary disjunctions of literals and m = min{u, w}. We use the notation [P,⋯,Q] ⇒ [R,⋯, S], where P, Q,⋯ are weighted clauses. It means that if there are some weighted clauses matching with [R,⋯, Q] (left side). they can be replaced by [R,⋯, S] (right side). We define the neighborhod resolution rule (NRES) as RES restricted to the A = B case. We also present the novel weighted modus ponens rule (MP) as: (x ∨ y, u), (x̄, w) ⇒ {(y, m) (x ∨ y, u Θ m) x̄, w Θ m) (x̄, ∨ ȳ, m) where m = min(u, w}. It is important to realize that this rule can be obtained by replacing B = false and y = A in the generic resolution rule (RES). Finally, we are studying the relation between the inference rules and the local consistency properties. For example, given an extension of the DPLL algorihtm for Max-SAT if it applies NRES0 rule at each node of the search tree then it enforces the NC* property. If DPLL applies both NRES0 and NRES1 at each node, it enforces AC*. NRESk denote NRES restricted to |A| = k with k >= 0. We present the equivalence of the
In this paper, we present a constraint-partitioning approach for finding local optimal solutions of large-scale mixed-integer nonlinear programming problems (MINLPs). Based on our observation that MINLPs in many engin...
详细信息
Structural decomposition methods have been proposed for identifying tractable constraint Satisfaction Problems (CSPs) [1-5]. the basic principle is to decompose a CSP into tree-structured sub-problems. the subproblems...
详细信息
暂无评论