Prediction of Solid Propellant Characteristic Signal Based on Improved BP Network

Jiu-ling ZHAO, He-hao SHANG, Wen-hai ZHANG

Abstract


The hidden layer node self-generating and adjusting method combining the improved BP neural network model is used to predict the solid propellant characteristic signal, the simulation result declares this method can overcome the "over-fitting" shortcoming.

Keywords


Genetic algorithm, Reject method, Penalty function method, Formulation optimization model


DOI
10.12783/dtcse/mso2018/20513

Refbacks

  • There are currently no refbacks.