Availability of Key Process Variable Estimates: An Evolutionary Approach

M. CARNERO, J. HERNANDEZ, M. SANCHEZ, A. BANDONI

Lisboa, Portugal

Congreso; ESCAPE 14; 2004

In this work a novel hybrid procedure is presented to solve the optimal design of a sensor structure, that satisfies both economic criteria and specific requirements on key variable estimates. The instrumentation network cost during its life cycle is minimized, subject to precision and availability constraints on variable estimates, for linear process. An Evolutionary Algorithm (EA) based on Genetic Algorithms (GA) with structured-population is developed and its performance is enhanced by local search. Application results are presented for a simplified plant of hydrodealkylation of toluene