A Stochastic Model Predictive Controller Based on Combined Conditions of Air Conditioning System for Electric Vehicles

He HongWen, Wang Chen, Jia Hui

Abstract


A stochastic model predictive controller based on combined conditions of air conditioning system is proposed to improve the energy efficiency of electric vehicles. According to the 25 groups of on-road tests, the velocity, solar radiation and temperature of driving conditions are recorded to construct combined conditions. A Markov chain-based combined conditions predictor is adopted to provide a prediction of the future conditions. In each predicted horizon, dynamic programming algorithm is applied to determine the best control strategy to balance the energy consumption and the cooling effect. The differences of three control strategies are compared in energy consumption and cooling effect, which are (i) the proposed SMPC method, (ii)a rule-based bang-bang controller and (iii) dynamicprogramming as the benchmark. Comparison resultsillustrate that, SMPC as an online method, theperformance is close to DP and is able to improve theAC energy economy by 7.86% than rule-basedcontroller.

Keywords


air conditioning, composed conditions, stochastic model predictive control, energy efficiency, thermal comfort


DOI
10.12783/dteees/iceee2018/27780

Full Text:

PDF

Refbacks

  • There are currently no refbacks.