Gross error detection and data reconciliation are importantproblems in operating chemical plants. Typically, constrained nonlinearoptimization techniques combined with statistical methods are used tosolve these problems. In this study, we explore the use of stochasticsearch for these purposes. One significant advantage of the method isthat it does not depend on any model structure information and onlyneeds simple algebraic calculation. Therefore, it is especially suitablefor gross error detection and data reconciliation of complicatedconnected processes
展开▼