China
Africa
Explore chapters and articles related to this topic
Software Available in Public Domain and the Commercial Software
Published in Nirupam Chakraborti, Data-Driven Evolutionary Modeling in Materials Technology, 2023
modeFRONTIER was developed by the Italian software company Esteco (Mode, 2021). It is now widely used, and has seen numerous real-life and industrial applications in recent years (Govindan et al., 2010; Russo et al., 2012; Carriglio et al., 2014; Jha et al., 2014). modeFRONTIER is meant for constructing data-driven models, as well as for multi-objective optimization. The users are allowed to select their own data source and they can use their preferred methods for modeling and optimization, for which various options are available. As Poles et al. (2008) discussed, modeFRONTIER, among other paradigms, supports Non-dominated Sorting Genetic Algorithm (NSGA-II), Multi-objective Game Theory, Evolutionary Strategies Methodologies, Normal Boundary Intersection (NBI) (Das and Dennis, 1998), among others. It also supports both neural net and genetic programming. In some of the studies (Govindan et al., 2010, Agarwal et al., 2009, Biswas et al., 2011), it was found that the neural net module in modeFRONTIER showed some tendency of overfitting. If any such problem happens, it is often difficult to fix, as the source code of this software is not available to users; the user interacts through a GUI. The users decide on the sequence of computing by constructing a simple flow chart on the screen. This allows combining different types of methods; for example, to combine an evolutionary training with a gradient-based optimization, modeFRONTIER can be quite efficiently used. Since several different types of modules are built in this software, the options are also ample. Interfacing with many other commercial software programs, for example, interfacing with prominent CAD (computer-aided design), CFD (computational fluid dynamics) software, is very much possible. It also allows interfaces for Excel, Matlab, and Simulink, and, as mentioned by Poles et al. (2008), parallel computing using modeFRONTIER is also possible. The software also provides extensive graphics support for the post-processing information.
Dynamic parameters optimization of straddle-type monorail vehicles based multiobjective collaborative optimization algorithm
Published in Vehicle System Dynamics, 2020
Junchao Zhou, Zixue Du, Zhen Yang, Zhouzhou Xu
The optimisation software modeFRONTIER integrated with the pre-processing software and the dynamic computing software is utilised for iteration, to reduce the iteration times and increase the computing efficiency. The random model is adopted in this article to replace the time-consuming dynamic computing process. The random variable sequence method adopted in this research has characteristics of randomness and irregularity. The random variable serial method is adopted to acquire the approximate model, and the multiobjective collaborative optimisation process of straddle-type vehicles with single-axle bogies is as shown in Figure 4. The flow of multiobjective collaborative optimisation is illustrated in Figure 5.