(626c) Cape-Open Compliant Multi-Objective Optimization Capability for Apecs Systems

Stringent environmental regulations and the necessity for economic profitability have placed numerous constraints on advanced power system plant designs. The process should have low pollutant emissions as well as cost competitive with existing technologies. Moreover some specialized objectives specific to the problem also need to be considered. Hence, designing advanced power systems involve consideration of a number of objectives. There are multiple objectives to be optimized simultaneously and the objectives are often conflicting and incommensurable. In such cases, a multi-objective optimization framework is utilized to compute the optimal objectives. The solution of a multi-objective optimization problem is not a single solution but a complete non-dominated or Pareto set [1], which includes the alternatives representing potential compromise solutions among the objectives. This makes a range of choice available to decision makers and provides them with the trade-off information among the multiple objectives effectively. A new non-linear optimization algorithm named ?MINSOOP? (Minimization of no. of Single Objective Optimization Problems) [2] based on the general framework of the constraint method and utilizing the efficienct Hammersley Sequence Sampling technique [3] has been developed. This algorithm requires solving significantly fewer number of single objective optimization problems than with both the traditional constraint method and the method based on the Monte Carlo sampling (MCS) technique to converge to the "true" mean and variance of the Pareto surface. The new algorithm provides tremendous computational savings compared to the traditional constraint method in obtaining an accurate representation of the Pareto set. To make this algorithm more accessible to users, we have created a CAPE-OPEN (CO) compliant tool which simplifies the process of setting up a multi-objective optimization framework for any process modeled in the APECS (Advanced Process Engineering Co-Simulator) to a great extent and guides the user in each step of the process through user-friendly graphical user interface (GUI) windows. APECS is an integrated software suite that combines the power of process simulation with high-fidelity, computational fluid dynamics (CFD) for improved design, analysis, and optimization of process engineering systems [4]. The APECS system uses commercial process simulation (e.g., Aspen Plus) and CFD (e.g., FLUENT) software integrated with the process-industry standard CAPE-OPEN (CO) interfaces. The CO standards provide a set of interfaces which allows the seamless integration of Computer Aided Process Engineering (CAPE) modules from various sources (software and equipment vendors, universities, and company generated) into process simulation environments [5]. This enables a process engineer to ?assemble' the necessary computational tools with the minimum effort from a collection of software (in-house, commercial, and/or academic) to achieve a best-in-class solution to various CAPE-related problems. The tool is built as a .exe file which can be installed in any computer running Windows OS. Once the tool is installed, it appears as an icon in the Aspen Plus CAPE-OPEN model library and can be dragged and dropped onto any flowsheet for multi-objective optimization. This tool automates the process of setting up a multi-objective optimization problem to a great extent and guides the user in each step of the process through user-friendly graphical user interface (GUI) windows. As a broad overview of the tool, there are three stages. The first stage is to specify the initial input variables which are:

? No. of objectives ? No. of decision variables ? No. of inequality constraints ? No. of equality constraints ? No. of samples for parametric RHS ? Min/Max specification for each objective

The next stage in the MOP setup is to import the Aspen blocks automatically created by the tool in the current directory. The final stage involves completing the minimal user additions required and running the simulation.The output file contains the values of the decision variables and objectives for each run. The tool was developed using Visual Basic 6.0 and FORTRAN. A case study of a advanced power system multi-objective optimization illustrates this capability.

References:

1. Diwekar, U. M.; ?Introduction to applied optimization', Kluwer Academic Publishers, Netherlands, 2003

2. Fu, Y and Diwekar, U.M.; ?An Efficient Sampling Approach to Multi-objective Optimization' Annals of Operations Research, 132, 109-134, 2004

3. Kalagnanam, J. R. and Diwekar, U.M.; ?An Efficient Sampling Technique for Off-line Quality Control', Technometrics, 39(3), 308-319, 1997

4. Zitney, S.E. et.al.; ?Advanced Process Engineering Co-Simulation of Power Generation Systems?, 30th International Technical Conference on Coal, Utilization & Fuel Systems Clearwater Coal Conference, Clearwater, FL, 2005

5. Koller, J. and Tobermann J.C; ?Global CAPE-OPEN ? D822 Migration Cookbook?, October 2002