A Movable Component Signal Identification Technology for Loose Particles Detection Based on Pulse Trajectory Tracking Algorithm

Abstract

The loose particle detection is an important part of the reliability detection of sealed electronic components in the aerospace field. Due to the complexity of the internal structure of aerospace sealed electronic components, the loose particles detection signals are often the aliasing of multiple types of signals, which makes it difficult to identify the detection results, especially the movable component signals. To solve this problem, we propose a new method for signal identification of movable components. First, the detection signal is obtained through the particle impact noise detection platform, the pulse sequence in the signal is extracted by energy and peak value threshold algorithm. Secondly, according to the vibration frequency and sampling rate, the position sequence of the dividing lines of the framing windows is obtained. Then, according to the framing window adaptive scaling algorithm, the continuity of the pulse is kept to the greatest extent and the influence of periodic instability is weakened through two scaling transformations. Finally, the pulse energy centroid trajectory is constructed in the framing window array to realize the fast identification of component signals. Experimental analysis based on typical examples shows that the proposed identification technology can provide an effective reference for the movable components signal identification of aerospace sealed electronic components. Theoretically, this method can be extended to other fields involving signal periodic instability problems, so it has extremely important reference and application value.