Research on Model Prediction Energy Management Strategy with Variable Horizon

Jianfei Cao; Jiankun Peng; Hongwen He

doi:10.1016/j.egypro.2017.03.819

Research on Model Prediction Energy Management Strategy with Variable Horizon

Jianfei Cao, Jiankun Peng, Hongwen He^*

^*Corresponding author for this work

School of Mechanical Engineering

Beijing Institute of Technology

Research output: Contribution to journal › Conference article › peer-review

9 Citations (Scopus)

Abstract

In order to improve the accuracy of model prediction energy management strategy, reconstruction and filling methods for state-loss of driving cycle are mainly studied. In this paper, the way to control the forecast accuracy by changing prediction time-scale is proposed and a real-time prediction model with variable horizon is constructed. Combined with dynamic programming, an energy management strategy based on model prediction control(MPC) with variable horizon is finally established. The correctness of this strategy is verified by hardware-in-loop(HIL) experiment. And the result show that the prediction accuracy could reach 8.203 km/h and fuel consumption is 18.3485L/100 km and the electricity consumption is 13.1081Wh, which has been improved comparing with traditional MPC with fixed horizon.

Original language	English
Pages (from-to)	3565-3570
Number of pages	6
Journal	Energy Procedia
Volume	105
DOIs	https://doi.org/10.1016/j.egypro.2017.03.819
Publication status	Published - 2017
Event	8th International Conference on Applied Energy, ICAE 2016 - Beijing, China Duration: 8 Oct 2016 → 11 Oct 2016

Keywords

HIL experiments
Model prediction energy management strategy
dynamic programming
variable horizon

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1016/j.egypro.2017.03.819

Cite this

@article{7bd75e50b5d3419a93e178d31a54f42c,

title = "Research on Model Prediction Energy Management Strategy with Variable Horizon",

abstract = "In order to improve the accuracy of model prediction energy management strategy, reconstruction and filling methods for state-loss of driving cycle are mainly studied. In this paper, the way to control the forecast accuracy by changing prediction time-scale is proposed and a real-time prediction model with variable horizon is constructed. Combined with dynamic programming, an energy management strategy based on model prediction control(MPC) with variable horizon is finally established. The correctness of this strategy is verified by hardware-in-loop(HIL) experiment. And the result show that the prediction accuracy could reach 8.203 km/h and fuel consumption is 18.3485L/100 km and the electricity consumption is 13.1081Wh, which has been improved comparing with traditional MPC with fixed horizon.",

keywords = "HIL experiments, Model prediction energy management strategy, dynamic programming, variable horizon",

author = "Jianfei Cao and Jiankun Peng and Hongwen He",

note = "Publisher Copyright: {\textcopyright} 2017 Published by Elsevier Ltd.; 8th International Conference on Applied Energy, ICAE 2016 ; Conference date: 08-10-2016 Through 11-10-2016",

year = "2017",

doi = "10.1016/j.egypro.2017.03.819",

language = "English",

volume = "105",

pages = "3565--3570",

journal = "Energy Procedia",

issn = "1876-6102",

publisher = "Elsevier Ltd.",

}

TY - JOUR

T1 - Research on Model Prediction Energy Management Strategy with Variable Horizon

AU - Cao, Jianfei

AU - Peng, Jiankun

AU - He, Hongwen

PY - 2017

Y1 - 2017

N2 - In order to improve the accuracy of model prediction energy management strategy, reconstruction and filling methods for state-loss of driving cycle are mainly studied. In this paper, the way to control the forecast accuracy by changing prediction time-scale is proposed and a real-time prediction model with variable horizon is constructed. Combined with dynamic programming, an energy management strategy based on model prediction control(MPC) with variable horizon is finally established. The correctness of this strategy is verified by hardware-in-loop(HIL) experiment. And the result show that the prediction accuracy could reach 8.203 km/h and fuel consumption is 18.3485L/100 km and the electricity consumption is 13.1081Wh, which has been improved comparing with traditional MPC with fixed horizon.

AB - In order to improve the accuracy of model prediction energy management strategy, reconstruction and filling methods for state-loss of driving cycle are mainly studied. In this paper, the way to control the forecast accuracy by changing prediction time-scale is proposed and a real-time prediction model with variable horizon is constructed. Combined with dynamic programming, an energy management strategy based on model prediction control(MPC) with variable horizon is finally established. The correctness of this strategy is verified by hardware-in-loop(HIL) experiment. And the result show that the prediction accuracy could reach 8.203 km/h and fuel consumption is 18.3485L/100 km and the electricity consumption is 13.1081Wh, which has been improved comparing with traditional MPC with fixed horizon.

KW - HIL experiments

KW - Model prediction energy management strategy

KW - dynamic programming

KW - variable horizon

UR - http://www.scopus.com/inward/record.url?scp=85020717342&partnerID=8YFLogxK

U2 - 10.1016/j.egypro.2017.03.819

DO - 10.1016/j.egypro.2017.03.819

M3 - Conference article

AN - SCOPUS:85020717342

SN - 1876-6102

VL - 105

SP - 3565

EP - 3570

JO - Energy Procedia

JF - Energy Procedia

T2 - 8th International Conference on Applied Energy, ICAE 2016

Y2 - 8 October 2016 through 11 October 2016

ER -

Research on Model Prediction Energy Management Strategy with Variable Horizon

Abstract

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this