Design and Application of Aerospace Accelerometer Testing System using Gap Statistics Based K-Means Clustering Method

In an aerospace industry, pyroshock testing is a crucial phase in developing the space electronics. With respect to fulfill a vibration exposure requires in maximum temperature surroundings like aerospace, maximum temperature manufacture as well as nuclear power, the fiber accelerometer is tested. The effectiveness of the error stability accelerometers is important for the inertial navigation system. The conventional overhanging accelerometers basically involves the flexible association structure, which constraints the long-term error stability. Hence, this research proposes the Gap Statistics-based K means clustering Method (GS-KCM) for the testing system for the aerospace accelerometer. GS supports to identify the optimal number of clusters through estimating the clustering output of K-Means with that of a random reference distribution. This make sures that the efficient number of clusters is selected without a need of manual selection of the data distribution. The proposed GS-KCM attains the better Mean Squared Error (MSE) of 1.293 and Mean Absolute Error (MAE) of 1.483 as compared to the conventional approaches like KCM and K-Nearest Neighbour (KNN).