In our research work to be presented, ceramic materials are designed with a multi-objective evolutionary algorithm which have optimal permittivity, an important property in the design of dielectric ceramics for microwave and telecommunications applications. Given the difficulty of large-scale manufacture and characterisation of these materials, we have attempted to guide a search for novel materials through the application of data-mining algorithms to a generic materials database which has been populated with data gleaned from the literature and contains composition and property information of over 1000 compounds. The composition-function relationship was then encapsulated using an artificial neural network.

Various constraints and objectives were incorporated to help search for designs that are optimal with respect to the permittivity, are stoichiometrically correct and are manufacturable. The stoichiometry is determined through charge calculations while a 'reliability index' calculation is used to ensure that permittivity predictions are likely to be accurate. In this way, we attempt to discover new compounds that are similar to the existing compounds stored in the database, along with new compounds that are very different. Finally, ceramic compounds that consist of 4 to 6 elements are given preference.

The use of a multi-objective design approach is particularly powerful since, rather than discovering a single ‘perfect’ solution, a range of solutions is found, where each solution satisfies the objectives and constraints to varying degrees. The final decision of ‘optimality’ is then left to a human decision maker, since this often involves other factors not originally encapsulated within the objectives or constraints. The algorithm used, along with the genotype (ceramic compound representation), constraints and objectives will be discussed, along with optimal compounds discovered by the genetic algorithm.

• Legal notice:
  

  Copyright (c) Pielaszek Research, all rights reserved.
  The above materials, including auxiliary resources, are subject to Publisher's copyright and the Author(s) intellectual rights. Without limiting Author(s) rights under respective Copyright Transfer Agreement, no part of the above documents may be reproduced without the express written permission of Pielaszek Research, the Publisher. Express permission from the Author(s) is required to use the above materials for academic purposes, such as lectures or scientific presentations.
  In every case, proper references including Author(s) name(s) and URL of this webpage: https://science24.com/paper/11229 must be provided.