Automotive transient management
The present diploma work proposes a cost-efficient solution for today’s current striking problem in automotive industry: the threat of microcontrollers being reseted. The problem and the resulting threat stems from the rapidly increasing number of electrical loads used in vehicles, and therefore causing the battery to be subjected to an ever-growing power demand. Furthermore, the currents of these loads are not constant over time but are subjected to transients, which could be even up to ten times higher than the static currents. These high peak current-transients could add up and cause – on account of the battery’s finite power capacity – voltage drops along the car’s supply line, on which both the loads and the control units are connected. If the voltage drops under 7.5 V, then the voltage-regulators can not provide the stable 5 V for the electronics and this could lead to the reset of the microcontrollers.
The spread of the start-stop technology highlights the justification of the topic, because there are more and more situations in which loads remain switched on beside stationary engine, requiring the battery not just to complement the alternator but to supply the loads’ aggregate power demand on its own. In the case of the low charged batteries voltage drops occur more often.
In contrast with the hardware based electrical disturbance suppression or the two-battery based solutions applied up to now, I propose a cost-efficient solution which – with the supplement of the software of ECU-s, that controls high-performance loads – makes the on-off switching of the loads schedulable, resulting in shifting the current-transients relative to each other in time and so preventing the formation of dangerous degree voltage drops. The modified banker’s algorithm, presented in this diploma work, uses information concerning the current and predictable state of the battery, gained by monitoring and modeling, and manages the resources of the battery shared among the loads, so it is used more effectively but does not disturb the operation of the safety-critical loads.
In this diploma work first I present the automotive loads and their transients based on literature data and measurements carried out by myself, then I analyze the supply line of the passenger cars. In an experiment I demonstrate that the degree of voltage drops could be minimized by scheduling. Finally using all of this information I plan the scheduling algorithm and the associated communication protocol.