Multi-objective Optimisation of Dynamic Responses for a Rail Freight Wagon using Regression Models

The optimization problem of the carbody dynamic response for a freight wagon fitted with three-piece bogie can be formulated as a multi-objective optimisation problem wherein four of the dynamic response parameters i.e. vertical acceleration on straight track in empty and loaded condition, lateral acceleration on 2° curve in empty and loaded condition, may be selected as representative objective functions for the overall dynamic response of the freight wagon. In this paper, attempts are made to form non-linear regression equations with experimental data to formulate the objective functions. After that, computational intelligence based evolutionary multi-objective optimisation is used to solve the problem and Pareto fronts are drawn for the objective functions using NSGA-II. Subsequently, the weighted optimization problem is solved for a different combination of weights.