ICT2014 Abstracts

A dc-dc converter with maximum power point tracking (MPPT) control is necessary for a wide output voltage ranged thermoelectric generator, which converts a higher voltage from thermoelectric generator down to the lower voltage needed to charge batteries in this study. To improve the tracking accuracy and speed, a novel MPPT control characterized by the aggregated dichotomy and gradient method is proposed. In the first stage, the dichotomy method is adopted as a fast search method for finding the approximate region of the maximum power. And then, the gradient method is applied for searching the point of maximum power quickly. In order to validate the proposed MPPT algorithm, a test bench of automobile exhaust thermoelectric generator (AETEG) is established for harvesting the automotive exhaust heat energy, which includes a 2.0 PSA 10LH3X engine, a self-designed thermoelectric generator, a dc-dc converter and an electronic load. The actual maximum power of AETEG at different engine revolution speed is depicted through the electronic load firstly. The MPPT control with the dichotomy method and the proposed method in this paper is embedded in the dc-dc converter, respectively. The experimental results of searching the maximum power through the dc-dc converter show that it spends 220ms searching the maximum power at rated speed for a single dichotomy method and just 140ms for the proposed method. Furthermore, the maximum power error is about 1.7% for a single dichotomy method and just 1.1% for the aggregated method compared with the actual value from the test results of electronic load. It is concluded that the proposed aggregated method in this paper is superior to the single dichotomy method for the MPPT control, especially enhancing the tracking accuracy and speed. The combination of dichotomy and gradient method can be used in the design and operation analysis of AETEG.