دانلود مقاله ISI انگلیسی شماره 91071
ترجمه فارسی عنوان مقاله

پیش بینی کننده سرعت مارکف و عملکرد پایه شعاعی استراتژی مدیریت زمان انرژی مبتنی بر شبکه عصبی برای پلاگین وسایل نقلیه هیبریدی

عنوان انگلیسی
Markov velocity predictor and radial basis function neural network-based real-time energy management strategy for plug-in hybrid electric vehicles
کد مقاله سال انتشار تعداد صفحات مقاله انگلیسی
91071 2018 18 صفحه PDF
منبع

Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)

Journal : Energy, Volume 152, 1 June 2018, Pages 427-444

پیش نمایش مقاله
پیش نمایش مقاله  پیش بینی کننده سرعت مارکف و عملکرد پایه شعاعی استراتژی مدیریت زمان انرژی مبتنی بر شبکه عصبی برای پلاگین وسایل نقلیه هیبریدی

چکیده انگلیسی

Power management strategy of plug-in hybrid electric vehicle for real-time application is a major challenge as the driving pattern is unknown beforehand. In this work, an innovative real-time power management strategy framework is proposed, including short horizon driving pattern prediction, driving pattern recognition, parameter off-line optimisation, parameter on-line prediction modelling, and power management strategy real-time application. A group of characteristic parameters is used to recognise driving patterns and the engine and motor working points are optimised globally by distributed genetic algorithm off-line. The optimised results approximation model is built on the basis of a radial basis function-neural network. Based on a linear programming algorithm, the higher order Markov velocity predictor is designed to obtain the short-horizon driving conditions. Combining the optimisation results approximation model, the real-time power management strategy is proposed. The on-line optimisation power management strategy comparing to the rule-based is analysed and the MATLAB/Simulink/AVL Cruise co-simulation results demonstrate that the fuel economy of real-time power management strategy improved by 16.3%, 12.7%, and 9.1% in HWFET, LA92, and Japanese urban driving patterns, respectively, which is largely higher than with a traditional rule-based strategy and slightly lower than, or approximately equal to, the global optimisation strategy.