Development of a high reliability, laser weld attched planar fibre coupling technique for Silicon Photonic devices

我们提出开发一种人工神经网络（ANN）来预测新型模具板系统的聚合物复合材料性能。人工神经网络应该使我们能够确定目标配方和工艺条件，从而最大限度地减少实验和开发所需的数量，以定制我们的产品以满足客户的要求。这项工作将在基础材料中已有的材料属性数据集上进行，人工神经网络的性能将在盲数据集上进行演示。

We propose to develop an Artificial Neural Network (ANN) for the prediction of polymer composite material properties of a novel tooling board system. The ANN should enable us to identify target formulations and process conditions thereby minimising the amount of experimentation and development required to tailor our product to customer’s requirements. The work will be conducted on data sets of materials properties already held within Base Materials and the performance of the ANN will be demonstrated on blind data sets.