Multiobjective Optimisation and Active Control of Bogie Suspension

Doctoral thesis, 2016

Railways provide fast, safe, clean, and cheap transportation service. The cost efficiency in railway operations can be scrutinized from different perspectives. Here, passenger ride comfort, wheel/rail contact wear, and safety (in particular running stability, track shift force, and risk of derailment) are considered as objective functions introduced to evaluate the dynamics behaviour of railway vehicles. Running speed also plays a key role in cost efficiency of railway operations. Higher speeds shorten journey time and make railways more competitive with other types of transportation systems. However, this might increase wear and deteriorate ride comfort and safety. To improve the performance in railway operations advanced designs and technologies are developed during the past decades. Bogie primary and secondary suspension systems of high speed trains can significantly affect the dynamics behaviour of the vehicle. Such components might have conflicting effects on different objective functions. It is important to have the optimum performance of suspension components. In this regard, one of the ultimate goals of this thesis is to improve the vehicle performance from different points of views by studying passive and active suspension systems and using multiobjective optimisation techniques to meet conflicting design requirements. Computational cost is one of the main challenges in multidisciplinary design optimisation. The computational efforts for optimisation can be significantly mitigated by narrowing down the number of input design parameters. Here, an efficient global sensitivity analysis is carried out to identify those suspension components that have prominent influences on different objective functions. Based on the global sensitivity analysis results obtained two multiobjective optimisation problems are formulated and solved. First, multiobjective optimisation of bogie suspension components with respect to safety to improve running speed on curves. Second problem is to reduce wear and improve ride comfort when the vehicle is operating with the enhanced speeds. Consequently, the vehicle runs secure and faster with higher ride comfort and less wear by means of the two optimisation problems solved. The optimisations are carried out using the genetic algorithm. In the case of safety optimisation problem, semi active control strategies are also applied using magnetorheological dampers and the effects on the dynamics behaviour are explored. The robustness of the bogie suspension Pareto optimised solutions against uncertainties in the design parameters is also studied. Active control technology is one of the main targets of this thesis. In this regard, a robust controller is designed using the H∞ control technique to stabilize the wheel set motion and improve curving performance. The controller is robust against track irregularities. Finally, the actuator dynamics is considered and a compensation technique is applied to reduce the actuator’s time delay and improve the performance.