A novel strategy for power sources management in connected plug-in hybrid electric vehicles based on mobile edge computation framework

Abstract

This paper proposes a novel control framework and the corresponding strategy for power sources management in connected plug-in hybrid electric vehicles (cPHEVs). A mobile edge computation (MEC) based control framework is developed first, evolving the conventional on-board vehicle control unit (VCU) into the hierarchically asynchronous controller that is partly located in cloud. Elaborately contrastive analysis on the performance of processing capacity, communication frequency and communication delay manifests dramatic potential of the proposed framework in sustaining development of the cooperative control strategy for cPHEVs.On the basis of MEC based control framework, a specific cooperative strategy is constructed. The novel strategy accomplishes energy flow management between different power sources with incorporation of the active energy consumption plan and adaptive energy consumption management. The method to generate the reference battery state-of-charge (SOC) trajectories in energy consumption plan stage is emphatically investigated, fast outputting reference trajectories that are tightly close to results by global optimization methods. The estimation of distribution algorithm (EDA) is employed to output reference control policies under the specific terminal conditions assigned via the machine learning based method. Finally, simulation results highlight that the novel strategy attains superior performance in real-time application that is close to the offline global optimization solutions.

Efficient Management of Energy Consumption of Electric Vehicles Using Machine Learning—A Systematic and Comprehensive Survey

“…Additionally, estimating IMU yaw misalignment by fusing information from automotive onboard sensors and an adaptive Kalman filter can enhance the accuracy of ML models in capturing vehicle dynamics [177]. IoT-based datasets [44,47,56,64,81,85,95,[106][107][108]110,117,122,125,126,128,134,136,138,140,144,149,155,172] 24…”

“…Offline [23,32,35,38,45,47,49,51,55,57,59,73,[79][80][81][82][83]86,88,91,94,103,107,109,110,116,122,128,131,133,134,136,141,149,156,157 The building models perform the training offline, which means that the prediction model is constructed once in the training phase and cannot be updated with new data. Therefore, over time, when the model gets out-of-date and does not work perfectly in the way that it should, it becomes necessary to re-train the model with more or newer data and then update the system which includes the new model.…”

“…Adaptive Neuro Fuzzy Inference System (ANFIS) [19,118,162] Multilayer feed forward NN (MFFNN) [99,118,121] Back Propagation NN (BPNN) [95,96,100,104,107,115,117,128,129,134,142,159] Multilayer Perceptron (MLP) [54,71,84,108,112,127,138,141,145,181] NARX NN [97] Stacked Autoencoder (SAEs) [116] ST-ResNE [151] Feed Forward NN [86] Extreme Learning Machine (ELM) [155] Recurrent 1 Kernel LMS (KLMS) [106] Fixed budget quantized KLMS (QKLMS-FB) [106] Recursive Least Squares (RLS) RLS [106] 1 Kernel RLS (KRLS) [106] RLS Tracker KRLS-T [106] Fuzzy Logic (FL) FL [32,113,123,142] 6 Fuzzy Inference System (Mamdami) [131] Interval Type-2 Fuzzy System (IT2FS) [153] Differential Evolution And Grey Wolf Optimizer (DEGWO) [110] 1…”

A novel data-driven controller for plug-in hybrid electric vehicles with improved adaptabilities to driving environment

“…The high-quality nonlinear characterization of complex operation performance and easy implementation of artificial intelligence technologies strengthens the optimal solution searching in control process, showing massive potential in promoting energy management of PHEVs. The raised three groups of methods, in general, have been validated efficient under certain conditions; whereas, the state-of-art application performance in real time is not maximized due to the limitation of the adopted algorithms and optimization target (Zhang, Y. et al, 2020b). For rule based strategies, despite easy application capabilities, expert-knowledge based design alienates the simple logic inspired strategies from optima.…”

An optimal control strategy design for plug-in hybrid electric vehicles based on internet of vehicles

“…To this end, there have been numerous studies undertaken considering strategies for minimization of energy consumption in PHEVs. These methods can be divided into the following three categories according to the characteristics of different algorithms: rule based strategies [3,4], global optimization based strategies [5,6], and instantaneous optimization based strategies [7,8].…”