Kernelized Proportional Intensity Model for Repairable Systems Considering Piecewise Operating Conditions

2012 ◽  
Vol 61 (3) ◽  
pp. 618-624 ◽  
Author(s):  
Yuan Fuqing ◽  
U. Kumar
Keyword(s):  
Operating Conditions ◽  
Repairable Systems ◽  
Proportional Intensity ◽  
Proportional Intensity Model ◽  
Intensity Model

Assessment of the effectiveness of corrective maintenance of an oil pump using the proportional intensity model (PIM)

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Sidali Bacha ◽  
Ahmed Bellaouar ◽  
Jean-Paul Dron
Keyword(s):  
Repairable Systems ◽  
Corrective Maintenance ◽  
Content Type ◽  
The Real ◽  
Oil Pump ◽  
Wide Range ◽  
Proportional Intensity ◽  
Proportional Intensity Model ◽  
Petroleum Company ◽  
Intensity Model

PurposeComplex repairable systems (CRSs) are generally modeled by stochastic processes called “point processes.” These are generally summed up in the nonhomogeneous Poisson process (NHPP) and the renewal process (RP), which represent the minimum and maximum repair, respectively. However, the industrial environment affects systems in some way. This is why the main objective of this work is to model the CRS with a concept reflecting the real state of the system by incorporating an indicator in the form of covariate. This type of model, known as the proportional intensity model (PIM), will be analyzed with simulated failure data to understand the behavior of the failure process, and then it will be tested for real data from a petroleum company to evaluate the effectiveness of corrective actions carried out.Design/methodology/approachTo solve the partial repair modeling problem, the PIM was used by introducing, on the basis of the NHPP model, a multiplicative scaling factor, which reflects the degree of efficiency after each maintenance action. Several values of this multiplicative factor will be considered to generate data. Then, based on the reliability and maintenance history of 12-year pump's operation obtained from the SONATRACH Company (south industrial center (CIS), Hassi Messaoud, Algeria), the performance of the PIM will be judged and compared with the model of NHPP and RP in order to demonstrate its flexibility in modeling CRS. Using the maximum likelihood approach and relying on the Matlab software, the best fitting model should have the largest likelihood value.FindingsThe use of the PIM allows a better understanding of the physical situation of the system by allowing easy modeling to apply in practice. This is expressed by the value which, in this case, represents an improvement in the behavior of the system provided by a good quality of the corrective maintenance performed. This result is based on the hypothesis that modeling with the PIM can provide more clarification on the behavior of the system. It can indicate the effectiveness of the maintenance crew and guide managers to confirm or revise their maintenance policy.Originality/valueThe work intends to reflect the real situation in which the system operates. The originality of the work is to allow the consideration of covariates influencing the behavior of the system during its lifetime. The authors focused on modeling the degree of repair after each corrective maintenance performed on an oil pump. Since PIM does not require a specific reliability distribution to apply it, it allows a wide range of applications in the various industrial environments. Given the importance of this study, the PIM can be generalized for more covariates and working conditions.

Robustness evaluation of a semi-parametric proportional intensity model

1994 ◽  
Vol 44 (1) ◽  
pp. 103-109 ◽  
Author(s):  
Waseem M. Qureshi ◽  
Thomas L. Landers ◽  
Edward E. Gbur
Keyword(s):  
Proportional Intensity ◽  
Proportional Intensity Model ◽  
Intensity Model

A Simulation-Based Random Process Method for Time-Dependent Reliability of Dynamic Systems

2010 ◽  
Author(s):  
Amandeep Singh ◽  
Zissimos P. Mourelatos ◽  
Efstratios Nikolaidis
Keyword(s):  
Random Process ◽  
Operating Conditions ◽  
Time Dependent ◽  
Process Time ◽  
Cumulative Probability ◽  
Repairable Systems ◽  
First Passage ◽  
True Value ◽  
Warranty Costs ◽  
Short Time

Reliability is an important engineering requirement for consistently delivering acceptable product performance through time. As time progresses, a product may fail due to time-dependent operating conditions and material properties, and component degradation. The reliability degradation with time may significantly increase the lifecycle cost due to potential warranty costs, repairs and loss of market share. In this work, we consider the first-passage reliability, which accounts for the first time failure of non-repairable systems. Methods are available that provide an upper bound to the true reliability, but they may overestimate the true value considerably. This paper proposes a methodology to calculate the cumulative probability of failure (probability of first passage or upcrossing) of a dynamic system with random properties, driven by an ergodic input random process. Time series modeling is used to characterize the input random process based on data from a “short” time period (e.g. seconds) from only one sample function of the random process. Sample functions of the output random process are calculated for the same “short” time because it is usually impractical to perform the calculation for a “long” duration (e.g. hours). The proposed methodology calculates the time-dependent reliability, at a “long” time using an accurate “extrapolation” procedure of the failure rate. A representative example of a quarter car model subjected to a stochastic road excitation demonstrates the improved accuracy of the proposed method compared with available methods.

Sign in / Sign up

Export Citation Format

Share Document