Brennan, Robert WilliamZeng, Zhigang2018-11-222018-11-222018-11-21Zeng, Z. (2018). Transfer lines of multi-state geometric machines: modelling, performance analysis, and control (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca. doi:10.11575/PRISM/34521http://hdl.handle.net/1880/109191A serial stage line is a typical pattern in manufacturing systems, but its modelling, analysis, and control during transients remain mostly unexplored. This type of manufacturing system consists of multiple serial stages. In each stage, each of which may have multiple machines, the system allows multiple independent machines to work concurrently. A major problem associated with system scheduling and dynamic capacity planning in serial stage lines is the uncertainty inherent to part delivery services, which may disrupt the original plan and result in higher production costs. Another problem relates to handling a lack of synchronization while coupling the stages into a system. A multi-state geometric machine is employed to describe one-part completion probability for all machines in a stage. The different states enable uncertain processing time to be modelled probatilistic distribution. Within a framework of serial stage lines with multi-state geometric machines and finite virtual buffers, Markov chain models of two- and three-stage transfer lines are proposed to describe the cascade effects of upstream and downstream resources, respectively. Based on the Markov model, a recursion model of the transition matrix is considered to be a state space. A predictive control model is then built to obtain the optimal admission policy for short-term scheduling, which can guide production managers in their decision-making. The proposed methodology is presented through different study cases. This model, combined with empirical studies of historical data, answers decision-making questions like, “How should parts be dynamically scheduled throughout a workday to minimize time-related production costs based on a fixed workload?” and, “If the original plan is interrupted by an emergency order or extraordinary duration processing, how should the rest of a day be rescheduled to minimize time-related cost based on a fixed workload?”engUniversity of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission.Engineering--Mechanicaldoctoral thesis10.11575/PRISM/34521