Optimization

Ray completed his PhD under the tutelage of the world-reknown and "father of structural optimization" Professor Lucien Schmit at UCLA. As such, Ray learned a number of gradient-based and genetic-based algorithms for solving tough optimization and function minimization problems.

Ray has used this knowledge in his post-schooling work to write papers related to control augmented structural synthesis, combined active/passive structural and controller optimization, and optimized composite layups for specified service loads. These are fun problems to solve for Ray!

At his professional workplace many people come to Ray to solve tough problems. He has a number of algorithms that can solve deterministic or heuristic optimization problems that many people don't even know how to start to solve. And Ray loves to share this knowledge and the innwer workings of the algorithms with people so that they can "go forth and multiply".

Heuristics

There are a number of scientific and engineering problems that cannot be determinstically specified. There are enough uncertainties in parameters or constraints that good solid values cannot be pinned down. These "open ended" problems are fun to solve!

Ray has developed a number of heuristic and combinatoric solutions for such problems. A recent case involved the procurement of "constant force" springs to a certain specification. After procurement it was found that the constant force springs did not produce a constant force and, furthermore, varied wildly in their stiffness properties. Ray developed a combinatoric optimization algorithm that selected sets of springs to pair together to minimize the effects of the wildly varying spring characteristics. This saved an enormous amount of money and time associated with trying to procure another set of springs from a different vendor.

This combinatoric algorithm and other Monte Carlo algorithms that Ray has produced allows parameters and constraints to be uncertain. A solution is achieved that accounts for the problem uncertainty and a statistical solution and a statistical set of error bounds is produced that allows decision to be made.

Nonlinear

Despite Ray liking to say, "Structures are linear, people are nonlinear", structures sometimes are nonlinear. Nonlinear structural analysis is difficult to perform and always open to technical prodding. Ray has thrived at producing a number of nonlinear models/simulations that have stood up to the technical prodding.

Statistics

Ray has been working on gathering sports statistics and finding new ways to display the statistics so that conclusions can be drawn. To view some preliminary National Football League team statistics for the past 15 seasons, you can click here. The process is interactive and you can display year-by-year statistics for a selected team, team-by-team statistics, or other views. Or you can view the current season statistics updated every Tuesday morning here.

To view daily Major League baseball team statistics for the current season, you can click here. The process is interactive and you can display various team statistics that are automatically updated daily. I'll add some additional team statistics in the next few weeks.

To view daily National Hockey League team statistics for the current season, you can click here. The process is interactive and you can display various team statistics that are automatically updated daily. I'll add some additional derived team statistics in the next few weeks.