A Bayesian approach to reliability prediction for one-shot devices Mr. Jun Seop LEE, Chin Uk LEE, Byeong Min MUN, Suk Joo BAE, Zae Ill KIM and Byung-Tae RYU (Hanyang University, Agency for Defense Development)
Projectile such as rocket and missiles is one of most important object for military and space industry. In order to measure its reliability accurately, it require many samples and time which result burden for development of projectile and industry. The cost of projectile is such burden to perform the reliability test as well as its required time for reliability test. In order to overcome this problem, alternative method to measure it reliability with small samples. Our test material is pin puller which convert combustion energy from the propellant to mechanical energy in rocket or missiles. In order to measure its lifetime with small samples, we apply bayesian method with proportional hazard model to estimate the its reliability.In this paper, we propose a semi-parametric bayesian approach by cox proportional hazards model. Finally, we compare the result to with result of other models to decide which model would be better for estimation of reliability.
A Study on the Characteristic of Stator Winding Degradation Process and Its Life Estimation of Induction Motors Mr. Viet Hung NGUYEN, Danwei WANG, Jeevanand SESHADRINATH, Abhisek UKIL, Vinay Kumar JAISWAL, Nishant VERMA, Viswanathan VAIYAPURI and Sivakumar NADARAJAN (Nanyang Technological University, Rolls-Royce Singapore Pte. Ltd.)
This paper presents an analysis on the characteristics of the stator winding degradation process. A diagnostic and prognostic parameter, which is derived using the sequence component approach, is proposed. It is estimated using only the measurements of voltages and currents, which are easily obtained, and hence, the method can be implemented in real-time applications. A test is designed for generating stator winding inter-turn fault and accelerating it. The characteristic of the degradation process based on the prognostic parameter is discussed. The collected data is modelled and the remaining useful life (RUL) is estimated.
Determining an Acceleration Factor for a Metering Pump used in a Polyurethane Injection Machine Dr. Gi-Chun LEE, Byung-Oh CHOI, Tae-Jin SONG, Jong-Sik CHOI and Chae-Young SUH (Korea Institute of Machinery & Materials, DUT Korea Co., LTD)
A metering pump for polyurethane injection machine has employed in the urethane production mixed with Isocyanate which is plasticizer and Polyol which is the raw material of polyurethane. The metering pump has used the piston type hydraulic pump which is the high-pressure mixed type and produced the products mixed with the constant flow by mixing head device supplying the fluid flow through the pipelines. While supplying the flow to the devices, the role of the metering pump supplying the constant flow is important. The study focused on determining acceleration factors (AF), as well as the accelerated life test method because a life test with normal operating conditions takes more than five years to carry out. This research selected the stress factor accelerated the main failure modes of the polyurethane injection metering pump, and the adopted acceleration model is an inverse power law. After selecting the acceleration factor and the model, the acceleration test performed as the pressure of acceleration pressure is 21 MPa, the operating pressure of the metering pump as field operating condition surveyed with 15 MPa.
Hamiltonian Monte Carlo sampling for Bayesian Hierarchical Regression in Prognostics Mr. Lachlan ASTFALCK and Melinda HODKIEWICZ (The University of Western Australia)
Advances in computational speed have enabled the development of many Bayesian probabilistic models due to Markov-Chain-Monte-Carlo (MCMC) posterior sampling methods. These models includes Bayesian hierarchical regression methods, which use group level information to inform individual asset predictions. Hierarchical models are increasingly used for prognostics as they recognise that the parameter estimates for an individual asset may be rationally influenced by data from other similar assets. Larger and high dimensional datasets require more efficient sampling methods for calculations, than traditional MCMC techniques. Hamiltonian Monte Carlo (HMC) has been used across many fields to address high dimensional, sparse, or non-conjugate data. Due to the need to find the posterior derivative and the flexibility in the tuning parameters, HMC is often difficult to hand code. We investigate a probabilistic programming language, Stan, which allows the implementation of HMC sampling, with particular focus on Bayesian hierarchical models in prognostics. The benefits and limitations for HMC using Stan are explored and compared to the widely used Gibbs Sampler and Metropolis-Hastings (MH) algorithm. Results are demonstrated using three case studies on lithium-ion batteries. Stan reduced coding complexity and sampled from posterior distributions more efficiently than parameters sampled with the Metropolis-Hastings algorithm. HMC sampling became less efficient with increasing data-size and hierarchical complexity, due to high curvature in the posterior distribution. Stan was shown to be a robust language which allows for easier inference to be made in the Bayesian paradigm.