Dp 165 en
Algorithms for nonlinear data reconsiliation
Author: Vítek Tomáš
In order to improve the accuracy of process data, the data reconsiliation method can be used. Data reconsiliation takes advantage of physical and chemical balances which can be arranged to decrease the measurement errors. Data processed by data reconsiliation are in harmony with the physical model of technology which is further used for optimization of the process. This thesis claims to formulate the data reconsiliation task in case of the heat system - the steam generator. The final form of problem is to minimize quadratic function subject to bilinear constrains. From various possible solutions we chose the gradient projection method and the sequential quadratic programming method (SQP). These too methods are compared in terms of efficiency and robustness in solving the reconsiliation of steam generator’s data.