Ben Spivey - The University of Texas at Austin

APMonitor modeling language has significantly reduced the time required to develop a fuel cell plant model that is central to my PhD dissertation. The modeling language provided an intuitive interface to model a distributed parameter, differential-algebraic dynamic system containing an implicit set of nonlinear algebraic equations. Since the DAE index is greater than 1, this problem could not be solved using Matlab's DAE mass matrices. I was able to quickly test several solvers, APOPT, BPOPT, and IPOPT, that can be interfaced with APMonitor by changing a single parameter. The language has provided a wide variety of configuration parameters to provide control over convergence criteria, varying levels of output for model debugging, and solver modes. A major benefit of the program is inclusion of many advanced solution modes: steady-state and dynamic simulation, parameter identification, real-time optimization, nonlinear control (NLC), and moving horizon estimation (MHE). My only personal experience with similar software is AMPL. The software can be compared to AMPL and GAMS but is unique from these in that it accepts DAE models and performs time discretization automatically versus an algebraic modeling format. The modeling interface is more similar to Honeywell NLC, but it includes additional solution modes, such as MHE, besides nonlinear control. The software may also be compared to gProms; however, gProms is recommended primarily for simulation, and researchers from our lab have reported difficulty linking gProms with a control algorithm. APMonitor competes with similar software in different aspects, but it is unique in the breadth, ease of use, and availability as a freeware software. In addition, an online interface helps facilitate their product support. As a fairly new software platform, APMonitor has had fewer field applications than more established software, but given its benefits, I expect field experience to continue increasing.

Benjamin Spivey
PhD Candidate
Energy Systems Modeling and Control
Department of Chemical Engineering
University of Texas at Austin

Klaus Schittkowski - University of Bayreuth

Your modeling language seems to be very clear and intuitively usable. I like it very much. If you need more test examples for ODEs or DAEs based on chemical engieering, please download the system EASY-FIT from my home page. If you need more software, let me know (1D PDEs\PDAEs).

Prof. Klaus Schittkowski
Department of Computer Science
University of Bayreuth
D - 95440 Bayreuth, Germany

comments powered by Disqus