Automated Design using Neural Networks and Gradient Descent
This addresses efficiency and speed issues in automated design for engineering applications, though it is incremental as it builds on existing neural network and gradient descent methods.
The paper tackles automated design for engineering tasks by using neural networks to mimic fitness functions and gradient descent to optimize designs, demonstrating effectiveness with optimized heat sinks and airfoils that maximize lift-drag ratios.
We propose a novel method that makes use of deep neural networks and gradient decent to perform automated design on complex real world engineering tasks. Our approach works by training a neural network to mimic the fitness function of a design optimization task and then, using the differential nature of the neural network, perform gradient decent to maximize the fitness. We demonstrate this methods effectiveness by designing an optimized heat sink and both 2D and 3D airfoils that maximize the lift drag ratio under steady state flow conditions. We highlight that our method has two distinct benefits over other automated design approaches. First, evaluating the neural networks prediction of fitness can be orders of magnitude faster then simulating the system of interest. Second, using gradient decent allows the design space to be searched much more efficiently then other gradient free methods. These two strengths work together to overcome some of the current shortcomings of automated design.