Multimodal Hidden Markov Model-Based Approach for Tool Wear Monitoring

Abstract

In this paper, a novel multimodal hidden Markov model (HMM)-based approach is proposed for tool wear monitoring (TWM). The proposed approach improves the performance of a pre-existing HMM-based approach named physically segmented HMM with continuous output (PSHMCO) by using multiple PSHMCOs in parallel. In this multimodal approach, each PSHMCO captures and emphasizes on a different tool wear regiment. In this paper, three weighting schemes, namely, bounded hindsight, discounted hindsight, and semi-nonparametric hindsight, are proposed, and two switching strategies named soft and hard switching are introduced to combine the outputs from multiple modes into one. As an illustrative example, the proposed approach is applied to TWM in a computer numerically controlled milling machine. The performance of the multimodal approach with various weighting schemes and switching strategies is reported and compared with PSHMCO.

Multi-model diagnostics for various machining conditions: A similarity-based approach

“…Moreover, a windowed variant of PSHMCO proposed in [14] that helps to reduce the computational cost is briefed.…”

“…Hidden Markov model (HMM) and hidden semi-Markov model are used in [6], [12][13][14] to recognize different tool conditions and health states in various applications. In [6], a physically segmented hidden Markov model-based approach with continuous output (PSHMCO) is proposed for diagnostics and prognostics in machinery systems.…”

“…In [6], a physically segmented hidden Markov model-based approach with continuous output (PSHMCO) is proposed for diagnostics and prognostics in machinery systems. Later on, in [14], a multi-modal version of the PSHMCO is introduced that can capture multiple degradation regimes with the same operating conditions but different cutters. However, in these studies the developed models are trained and tested on the data from the same operating conditions, the experiments under study in this paper are from various operating conditions with multiple trials for each operating condition.…”

“…In Section II, a hidden Markov model-based approach from [6] and its windowed version [14] are reviewed as the core utilized further in the proposed approach. A similarity based multi-HMM diagnostics approach is proposed in Section III.…”

Fault Estimation and Fault Tolerant Control for Discrete-Time Dynamic Systems

“…Except for the hardware redundancy in some key components, information redundancy has gained more and more attention in both academic community and industries for the last four decades owing to the convenience for implementation and significant saving in the cost. The fruitful theoretic results produced by a variety of fault diagnosis methods such as model-based methods [1][2][3][4][5], signal based methods [6][7][8] and data-driven methods [9][10][11], and their applications in wind energy systems, robotic manipulators, power electronics, motor drive, power quality, Manuscript received June 21, 2014; revised November 6, 2014; accepted December 15, 2014. Copyright © 2014 IEEE.…”