Scientific Foundation
Institute for Operations Research and Computational Finance (ior/cf), University of St. Gallen, Switzerland (http://www.iorcf.unisg.ch/)
The ior/cf is leading in research and development of stochastic optimization and simulation models for the energy and finance industry. More than 10 years of experience in basic and applied research in the field of optimization under uncertainty are implemented in the solutions of the ior/cf. The ior/cf is headed by Prof. Dr. Karl Frauendorfer, an internationally renown scientist in the field of Stochastic Optimization.
The mathematical approaches to optimization of complex decisions problems are tailored to the specific requirements of clients. Experiences gained in the banking and finance industries are applied in models for power and gas utilities and contribute to increased efficiency and profit. ior/cf provides stochastic optimization models for power generation, trading and retail companies that operate in increasingly uncertain environments. It has a proven track record of enhanced decision support in long and short-term operations and trading in the power industry.
_________________________________________________________________________________
Mathematical Optimization Solver
ILOG (www.ilog.com)
At the heart of the DT@Energy optimization Suite ILOG CPLEX is embedded as the optimization kernel that solves the mathematical model. ILOG CPLEX is the best solution for solving complex mixed integer mathematical problems especially given the complexity of the mathematical model representing utilities complex trading and generation portfolios.
ILOG CPLEX is designed to solve corporate business problems, using the most sophisticated analytical techniques. Major companies and software providers in supply-chain planning, network design, transportation logistics, utilities and a variety of other industries depend on CPLEX in mission-critical resource allocation applications. CPLEX is at the core of countless cutting-edge commercial products worldwide.
ILOG Parallel CPLEX is a parallel computation engine for solving most difficult real-world industrial problems. Pioneering a technological breakthrough, ILOG is the first software vendor to develop commercial parallel algorithms in partnership with multiple parallel-computer vendors. Parallel algorithms allow you to solve previously unsolvable problems on a wide range of high-performance computers.
Speedier solutions, increased throughput
ILOG Parallel CPLEX provides a number of different options for improving performance. ILOG CPLEX Concurrent Optimizer can apply multiple algorithms to a single linear programming problem. Each algorithm operates on a different CPU. This technique results in the fastest solution time, without having to predetermine the best algorithm.
Another approach is to use multiple CPUs with a single algorithm. This technique solves a single linear program, or a mixed-integer program, much faster than a single CPU.
The mathematical models that are implemented within DT@Energy involve multiple binary variables, which define the model as mixed integer model. As opposed to plain linear programs, which are solved by conventional Simplex algorithm, mixed integer models need to be solved by branch and cut algorithms. This requires multiple solving of the model with different settings of the binary variables. These multiple jobs are predestinated to be distributed on a large number of compute servers, which is supported by Parallel CPLEX. The time for solving the overall model can be cut down to fractions of the time needed for a single computer.
ILOG CPLEX provides advanced methods for tuning the desired solution of the model with respect to calculation times. Various settings for pay offs between proven optimality ranges and calculation times are available. Depending on users' individual requirements in terms of calculation times different settings can be implemented.