Luis Gimeno Latre

Short-term planning and scheduling in multipurpose batch chemical plants: a multi-level approach

Abstract This paper deals with short-term planning/scheduling problems when the product demands are driven by customer orders. This type of problem is, in general difficult to be solved, once the orders are subject to changes in short periods. In this paper is proposed a multi-level decomposition procedure, containing at least two levels. At the planning level, demands are adjusted, a raw material delivery plan is defined and a capacity analysis is performed. Therefore, time windows for each operation are defined. At the scheduling level, an MILP model is developed. In order to reduce scheduling problem dimension, it is performed an analysis of equipment contention profile, enabling to model only situations where a decision variable is requested.

SEQUENTIAL APPROACH TO PRODUCTION PLANNING IN MULTISITE ENVIRONMENTS

Abstract A sequential planning approach is proposed for make-to-order situations that tries to emulate the practitioners approach. It deals sequentially with assignment of sites and transportation, completion time, exact timing and detailed sites planning, allowing negotiation among sites in terms of runtime required for each site.

An Interactive Optimization Technique for Short Term Scheduling of Batch Multipurpose Plants

Abstract In this paper is discussed a reactive approach for the scheduling of multipurpose chemical batch plants based on the utilization of a rolling time horizon and a lookahead procedure. It is supposed that a previous planning activity has determined product batches to be processed and its due dates in order to maximize the total value added in terms of products deliveries amounts. The proposed short term scheduling technique is started when perturbations in the plant occur and cannot be absorbed in the previous scheduling. The proposed technique is based on a MILP formulation of the State Task Network Representation.

Frequency Criteria Specification for Generalized Minimum Variance Controller

Abstract For a generalized minimum variance controller algorithm the weighting polynomials are calculated in a way to assign the closed-loop poles of the system and to specify the controller gain at a given frequency. The technique is based on the property that the pole assignment equation may have multiple solutions if the weighting polynomials are not of minimal order. By increasing the order of these polynomials, the degree of freedom obtained is utilized for controller gain adjustment in high fre quencies. As a result the oscillations in the control signal may be reduced without changing the deterministic behaviour of the system.