Planning and sequencing heuristics for feature-based control of holonic machining equipment

Abstract

The intelligent manufacturing system program was proposed by Japan in 1989. Five participating regions-Australia, Canada, the European Community, Japan, and the United States-currently are involved in developing 21st century manufacturing technology through an investment of US $1.2 billion over 10 years. Korea joined the program and will start work on one of the six ongoing projects, holonic manufacturing systems (HMSs). The objective of the paper is to develop the control architecture of the holonic machining unit (HMU) for construction of the HMSs and to present some planning and sequencing heuristics for feature-based control of the HMU. Further, the paper provides the HMU's functionality using the IDEF0 function modeling method. The basic operation of the decision maker among the HMU's functions is to determine an efficient feature sequence in real time from the nonlinear feature graph used to represent a process plan. To this end, two methodologies are applied sequentially to managing a nonlinear process plan: removal of the OR nodes and then grouping and sequencing the features in the feature graph. Markov chain theory is used to compute the path preference indicator for removing the OR nodes, that is, for selecting the best path among those surrounded by OR nodes. The resulting graph is the AND graph, from which the feature type nodes are formed into sequenced groups. The CNC codes associated with the features in each group are combined and downloaded to the CNC machine. The development of the methodologies can help manufacturers efficiently cope with unexpected failures encountered during computer-automated machining.