Analysis of Possibilities and Effectiveness of Combine Rough Sets and Neighborhood Theories for Solving Dynamic Scheduling Problem

In our paper we considered dynamic tasks scheduling problem. To solve this problem we try applying both characteristics of rough sets and neighborhood theories (theory name introduced for temporary use). Among rough sets characteristics we can quote proposal of Slovinski et al.: outranking parameter, reference ranking, cumulated preference etc. Among neighborhood characteristics we can appoint Jaron, Nikodem et al. terms such as subordinate, tolerance and collision. The tasks succeeding process has an asynchronous character. However, tasks assigning and loading to actual processor is done in regular time intervals. The tasks choice problem is supported by several approximate scheduling algorithm. Usually the complexity of this kind of algorithms is much lower than of precise algorithms. According to tasks placements appointed by every algorithm a dynamic data table is created. This data table provides information on estimations of both characteristics types, which support inference process of tasks sequencing in a form of final scheduling list. Using consolidated approach we simplify inference organization and obtain tools for taking into account dispersed tasks locations according to different algorithms.