Preliminary evaluation of Bollinger deadband filtering using upper and lower bands
Nunzio Marco Torrisi · 2014
In this paper, the author investigates the use of the Bollinger Bands™ theory to compute the deadband sampling algorithm of unbounded variables. In this scenario the dead-band algorithm has to be adapted to compute the Altering using a new dynamically preset interval derived from the Bollinger Bands theory for financial applications. The number of filtered samples and the integral absolute error are adopted as metrics to compare performance results of the new original deadband proposed, named Bollinger Deadband, with the results of the classic absolute deadband algorithm. The signals adopted for the simulation are based on the pseudo periodic time series. For these signals the effectiveness of the Bollinger deadband is explored as alternative sampling method when the signal limits are unknown a priori.