Dilek Alkaya

Carnegie Mellon University, Pittsburgh, Pennsylvania, United States

Are you Dilek Alkaya?

Claim your profile

Publications (3)2.24 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: This study describes a general framework for coupling and optimizing multiple models with multiplier-free reduced Hessian successive quadratic programming. This tailored approach enables the use of existing process simulators/models simultaneously with the optimizer, and this leads to efficient solution of process optimization problems. In this paper, a unified strategy is proposed to combine the model information and then decompose the sensitivity matrices that arise from the connections of the streams. With the proposed decomposition approach, the need for storing a large constraint matrix is avoided and individual models that only pass their Newton corrections and sensitivity information are left. The solution avoids problems due to model failure and can save a lot of time because models are not converged at intermediate optimization iterations. The resulting tailored optimization approach is illustrated on the dynamic optimization of a batch reactor/column system selected from Yi and Luyben (Comput. Chem. Eng. 1997, 12, 25).
    Industrial & Engineering Chemistry Research 04/2000; 39(6). · 2.24 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: This article discusses the use of SQP in industry, together with techniques for mathematical optimization and process modeling to improve economic performance of plants in process industries. First, different types of flowsheeting optimization problems based on SQP methods and process models are introduced briefly. Then a number of process optimization formulations and strategies are discussed, along with how the SQP algorithm needs to be developed and extended to take advantage of large-scale systems. In particular, the development of reduced Hessian SQP (rSQP) is presented along with different variants. Finally, literature on industrial and academical applications of SQP and rSQP is given.
    01/1999;
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: With the increasing size and complexity of process simulation and optimization problems, exploitation of the process model becomes increasin gly important. While most researchers recognize the need to exploit the probl em structure, for instance at the linear algebra level, this study explores the case when mu ltiple model solvers are required for simulation and optimization of the overall process system. While this approach is standard for process flowsheeting, we need to consider how w e can take advantage of sophisticated simultaneous solution and optimization strategies f or large-scale optimization. Here we discuss both open form and closed form models, and demonstrate that both are needed for different types of problems. We then consider an ap proach where closed form or 'black box' models can be 'opened up' to achieve simultane ous optimization without disturbing the inherent structure of the model's solver. In ad dition, several applications, including process flowsheets, dynamic optimization, PDE models and process integration are highlighted. Finally, we close with some challenges and areas for future work for both modeling environments and optimization algorithms.