Prediction-Based Fast Thermoelectric Generator Reconfiguration for Energy Harvesting from Vehicle Radiators

Introduction

Many researches have been focused on improving the efficiency of energy harvesting devices, among which thermoelectric generator (TEG) is a wide-used device that generates electric energy directly from heat energy via Seeback effect [1]. For most of the energy harvesting systems on vehicle, a number of TEG modules need to be combined together for extensive contact with heat sources. However, without a rational arrangement for the connections between modules, the system will suffer from a poor holistic performance due to electrical limits. Besides, temperature fluctuation during the harvesting process brings significant disturbance to the system. Thus a system-level solution is necessary for achieving a high conversion efficiency for each TEG module.

Radiator and Teg Module Modeling

Thermoelectric generation (TEG) has increasingly drawn attention for being environmentally friendly. A few researches have focused on improving TEG efficiency at system level on vehicle radiators. The most recent reconfiguration algorithm shows improvement on performance but suffers from major drawback on computational time and energy overhead, and non-scalability in terms of array size and processing frequency. We propose a novel TEG array reconfiguration algorithm that determines near-optimal configuration with an acceptable computational time. More precisely, with O(N) time complexity, our prediction-based fast TEG reconfiguration algorithm enables all modules to work at or near their maximum power points (MPP). Additionally, we incorporate prediction methods to further reduce the runtime and switching overhead during the reconfiguration process. Experimental results present 30% performance improvement, almost 100 reduction on switching overhead and 13enhancement on computational speed compared to the baseline and prior work. The scalability of our algorithm makes it applicable to larger scale systems such as industrial boilers and heat exchangers.

HeadShot

Experimental Results

Our prediction results presented in this work are under parameters which are proved to have the best performance. A 1-second prediction MAPE comparison of these three methods are shown in Fig. 5. MLR method has the best performance for temperature prediction on vehicle radiator. Even the highest percentage error of 2-second MLR prediction in this duration with such a radical temperature fluctuation is only around 0.3%. Due to the low time complexity (O(N)) of MLR, the temperature prediction process is so transitory that it lays ignorable affects on the reconfiguration algorithm’s runtime.

HeadShot