The term multi-objective optimization may sound some-what foreign, but I promise you, you've been using it yourself every day, for years. Bare with me..
Say you’re buying a car. You’ll probably want a performant and an attractive one, but at the lowest possible price. The choice criteria you’re working with, are: economic convinience, performance and aesthetics. Notice that if you choose to maximize performance in your choice, you’ll probably be minimizing economic convinience…Tricky one, right?
You may not have realized this, but your final decision will rely on you performing a “multi-objective” optimization. And the catch with multi-objectivity is that often the goals are conflicting, and there is more than one “optimal” solution, depending on which criteria are most important to you.
Moving back to the engineering field..
In the development process of technical products, like a car for example, there are usually several criteria that must be scientifically considered and met, like low emissions, low weight, high performance, low noise, low cost etc. An engineeer’s job is to find the design parameters that ensure the product meets these criteria.
In a traditional, experience-based approach to product development, the engineer would start by finding a design that maximizes one of the criteria, like low emission levels. Then he’d check that this design respects the desired product weight, and maybe he’d discover it’s too heavy. So he’d go back to modify the intial design, make it meet the weight criteria, only to find out that it doens’t attain the desired performance…
There’s never enough time to check on all the possible design-combination options, is there? And the engineer could easily be overlooking an even better solution that he just didn’t condsider. This is where multi-objective simulation software comes-in handy.
KIMEME is an innovative multi-objective optimization and multi-disciplinary platform for CAE applications, developed by Cyber Dyne
Within KIMEME’s optmization platform, the user can:
- set up the optimization problem, specifying the design parameters that need to be identified by Kimeme
- identify the objectives that should be pursued (maximization, minimization or containment), and the constraints that need to be respected
- select the CAE simulation software that are employed within the company to support the design process (CAD, 0D/1D, CFD-3D, FEM..) and create the workflow Kimeme should follow to compute the parameters.Kimeme will automatically integrate and execute the CAE simulation software that have been selected.
Kimeme assignes values to the free parameters and feeds them to the first CAE software in the workflow, collects the outputs and uses them as inputs for the next software in the workflow. Kimeme’s algorythms will analyze the results and sequentially adjuste the initial parameter values, testing them untill it obtains a range of “optimal” design paramters that satisfy the user’s objectives.
Kimeme also includes a set of data analysis tools to help the user review, understand, filter and organize outputs in a smart and efficient way.