Percorrer por data de Publicação, começado por "2022-05-09"
A mostrar 1 - 2 de 2
Resultados por página
Opções de ordenação
- Systematic method for developing reference driving cycles appropriate to electric L-category vehiclesPublication . Watling, David; Baptista, Patrícia; Duarte, Gonçalo; Gao, Jianbing; Chen, HaiboIncreasingly, demanding environmental standards reflect the need for improved energy efficiency and reduced externalities in the transportation sector. Reference driving cycles provide standard speed profiles against which future developments and innovations may be tested. In the paper, we develop such profiles for a class of electric L-category vehicles, which are anticipated to play an increasing future role in urban areas. While such driving cycles exist for regular L-category vehicles, these may not be suitable in the case of electric vehicles, due to their power output limitations. We present a methodology for deriving these new driving cycles, developed from empirically deduced power relationships, before demonstrating their application under different assumptions on the terrain and vehicle characteristics. The applications demonstrate the feasibility of the method in developing appropriate driving patterns for alternative real-world contexts. On flat terrain, the adjustments made to cope with the power limitations of L-EV do not introduce significant differences in energy consumption, suggesting that the certification does not require extensive modification. However, when considering road slope, differences of up to 5% in energy use and up to 10% in regenerated energy were observed, showing the importance of the developed method for assessing vehicle performance in real-world driving.
- Particle swarm optimization-based algorithm for optimal reactive power dispatchPublication . Menezes, Ruben Henrique de Brito Azancot de; Fonte, Pedro Miguel Neves da; Pestana, Rui José Oliveira NóbregaAbstract Reactive power optimization, more specifically the Optimal Reactive Power Dispatch (ORPD), is very important for the security and economy of power systems. It is a mixed integer non-linear optimization problem for which metaheuristic methods have proven to be effective in its solution. These methods do not guarantee that a global optimum is found, as premature convergence to local optima may occur. Much research is focused on improving the ability of metaheuristics in finding the global optimum. This thesis presents a particle swarm optimization-based algorithm (PSO) for the solution of the ORPD. The implementation of the PSO algorithm is explained in detail, including the parameter selection, a constraint handling method and a discrete variable rounding method. Another version of the PSO, called Fitness-Distance Ratio PSO (FDR-PSO), is implemented in order to improve the results of the basic PSO, by decreasing the chance of premature convergence to local optima. One other version of the PSO called Second Order PSO (SO-PSO) is implemented for the same purpose. The SO-PSO until this point remained somewhat untested in the ORPD. This second-order principle was combined with the FDR-PSO, making it the SO-FDR-PSO. The implemented versions of the PSO were tested on the IEEE 14 bus and IEEE 39 bus systems with 30 runs each. On the 14 bus system, the lowest active losses found were of 12,280 MW. The basic PSO had an average of 12,387 MW showing that it often converged to good local optimal solutions, and occasionally 12,280 MW, which is possibly the global optimum (the best optimal solution the PSO was able to find). The solutions found were also always feasible, meaning the ORPD was effectively solved. The FDR-PSO was able to improve the results, with an average of 12,297 MW. The SO-PSO improved the results further, the most notable improvement being in the success rate, now of 57%, which means it often found solutions close to 12,280 MW, the apparent global optimum of the problem, and was observed to never converge prematurely to local optima. The SO-FDR-PSO had an average of 12,280 MW and success rate, of 100%, meaning the apparent global optimum was consistently reached. On the 39 bus system the overall results worsened in statistical terms in relation to the 14 bus system, although the solutions remained always feasible, still with a considerable decrease in active losses and the FDR-PSO/SO-FDR-PSO maintained its superiority over the PSO/SO-PSO.