In this work, solution strategies for the optimal design of nonredundant observable linear sensor networks are discussed. The Greedy algorithm allows the problem only to be tackled for a subset of optimization criteria. Particular deterministic techniques or general evolutionary strategies are necessary to solve the problem for more complex objective functions. In this context, a procedure based on the application of genetic algorithms (GAs) and linear algebra is presented. Ad hoc operators are designed for the crossover and mutation operations because the classic genetic operators perform poorly. In contrast to ad hoc deterministic codes, which find the design solution for each specific criteria, this strategy allows the problem to be solved with different objective functions using the same implementation. Furthermore, this code is extended to handle multiobjective problems through a modification of only the selection operator. An industrial example is provided to show the efficiency of the algorithm.
In this work, a procedure for solving the optimal design and upgrade of linear sensor networks, subject to quality constraints on a set of key variable estimates, is presented. The strategy aims to select the optimal set of flowmeters without imposing restrictions on the mathematical nature of the objective function and constraints. An evolutionary technique based on genetic algorithms (GAs) is proposed that combines the benefits of using structured populations in the form of neighborhoods and a local search strategy. Both procedures take advantage of existing process knowledge. Application examples are provided for the instrumentation design of a steam metering network of a methanol plant, which show that the algorithm has a good balance among its exploration and exploitation capabilities.
This work addresses the cyclic scheduling of cracking furnace shutdowns in ethylene plants within a shortterm production planning model, based on a discrete time representation. Cracking furnaces are continuous parallel reactors that show decaying performance during their operation due to coke deposition on coil walls. For that reason, they must be periodically shutdown and cleaned. This behavior is modeled through binary variables and coil internal roughness, a variable whose increase has a linear dependence on operation time. After cleanup, roughness is at its lowest value and starts increasing again during operation. The cyclic scheduling model includes not only furnaces models but an entire plant mathematical model at each time interval to carry out production planning for meeting varying demands, as well as to determine main plant operating variable profiles and to predict an ethane recycle stream, which is an important feed to cracking furnaces and constitutes a key variable for the optimal shutdown schedule. The model includes nonlinear mathematical functions for each cracking furnace production as a function of main process variables, simplified models for distillation columns in the separation train, and raw material and product storage equations. Additional binary variables are included to force null values for production in shutdown furnaces. The resulting mixed-integer nonlinear programming (MINLP) model is solved in GAMS with DICOPT++.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.