The Experts below are selected from a list of 11385 Experts worldwide ranked by ideXlab platform

Frede Blaabjerg - One of the best experts on this subject based on the ideXlab platform.

  • System-Level Reliability Analysis of a Repairable Power Electronic-Based Power System Considering Non-Constant Failure Rates
    2020 22nd European Conference on Power Electronics and Applications (EPE'20 ECCE Europe), 2020
    Co-Authors: Amirali Davoodi, Yongheng Yang, Tomislav Dragičević, Frede Blaabjerg
    Abstract:

    Conventionally, for reliability studies in power systems, a Constant Failure Rate is assumed for power generation units. If these units consist of power electronic converters, this assumption might not be valid due to the aging of power components, and it will lead to an unrealistic prediction of reliability. On the other hand, at the system-level, commonly-used reliability calculation tools, such as Monte Carlo Simulation (MCS) and Continuous Markov Process (CMP), are either time-consuming or not possible to be applied to the systems with non-Constant Failure Rates. Therefore, in this paper, a methodology is proposed to calculate system-level reliability for a Power-Electronic-based Power System (PEPS), consisting of several converters with non-Constant Failure Rates. By doing so, not only is the effect of mission profiles integRated into the system-level reliability model, but also the wear-out Failures and corrective maintenance are considered. Finally, for a case study PEPS, the system-level indices are calculated using the proposed method. It is shown that assuming Constant Failure Rates for PEPS units is inaccuRate and misleading. Moreover, the impacts of various factors, e.g., mission profile, repair Rate, topology, and rating of converters, on the system-level reliability are investigated and analyzed.

  • Prediction and Validation of Wear-Out Reliability Metrics for Power Semiconductor Devices With Mission Profiles in Motor Drive Application
    IEEE Transactions on Power Electronics, 2018
    Co-Authors: Ui-min Choi, Frede Blaabjerg
    Abstract:

    Due to the continuous demands for highly reliable and cost-effective power conversion, quantified reliability performances of the power electronics converter are becoming emerging needs. The existing reliability predictions for the power electronics converter mainly focus on the metrics of lifetime, accumulated damage, Constant Failure Rate, or mean time to Failure. Nevertheless, the time-varying and probability-distributed characteristics of the reliability are rarely involved. Moreover, in the public literatures, there are few evidences showing that the accuracy of the predicted reliability was experimentally validated. In this paper, a more advanced metric “cumulative distribution function (CDF)” is introduced to predict the reliability performance of the power electronics system based on mission profiles in motor drive application. Furthermore, the accuracy of the predicted reliability metrics is verified through a series of wear-out tests in a converter testing system. It is concluded that the CDF is a very suitable metric to predict the reliability performance of the converter, and it has shown good accuracy with much more reliability information compared to the existing approaches. In this method, the correct stress translation and dedicated strength tests based on mission profiles are two key factors to ensure the efficiency and accuracy of reliability prediction.

Huai Wang - One of the best experts on this subject based on the ideXlab platform.

  • mission profile based system level reliability analysis of dc dc converters for a backup power application
    IEEE Transactions on Power Electronics, 2018
    Co-Authors: Dao Zhou, Huai Wang
    Abstract:

    Reliability analysis is an important tool for assisting the design phase of a power electronic converter to fulfill its life-cycle specifications. Existing converter-level reliability analysis methods have two major limitations: 1) being based on Constant Failure Rate models; and 2) lack of consideration of long-term operation conditions (i.e., mission profile). Although various studies have been presented on power electronic component-level lifetime prediction based on wear-out Failure mechanisms and mission profile, it is still a challenge to apply the same method to the reliability analysis of converters with multiple components. Component lifetime prediction based on associated models provides only a $B_{X}$ lifetime information (i.e., the time when X % items fail), but the time-dependent reliability curve is still not available. In this paper, a converter-level reliability analysis approach is proposed based on time-dependent Failure Rate models and long-term mission profiles. Two different methods to obtain the component-level time-to-Failure are illustRated by a case study of dc/dc converters for a 5 kW fuel cell-based backup power system. The reliability analysis of the converters with and without redundancy is also performed to assist the decision making in the design phase of the fuel cell power conditioning stage.

  • The Impact of Topology and Mission Profile on the Reliability of Boost-type Converters in PV Applications
    2018 IEEE 19th Workshop on Control and Modeling for Power Electronics (COMPEL), 2018
    Co-Authors: Saeed Peyghami, Pooya Davari, Huai Wang
    Abstract:

    This paper investigates the impact of different converter topologies and mission profiles on the reliability of dc/dc boost-type PV converters. The reliability of three boost-type converters with the same input/output specifications is modeled employing a mission profile-based reliability evaluation method considering non-Constant Failure Rate for the electrical components. This study identifies the contribution of active and passive components on the converter reliability for identifying the most Failure prone components. Furthermore, the applicability of converter structures for different climate conditions is demonstRated seen from the reliability point of view.

Zizhong Chen - One of the best experts on this subject based on the ideXlab platform.

  • fail stop Failure algorithm based fault tolerance for cholesky decomposition
    IEEE Transactions on Parallel and Distributed Systems, 2015
    Co-Authors: Doug Hakkarinen, Zizhong Chen
    Abstract:

    Cholesky decomposition is a widely used algorithm to solve linear equations with symmetric and positive definite coefficient matrix. With large matrices, this often will be performed on high performance supercomputers with a large number of processors. Assuming a Constant Failure Rate per processor, the probability of a Failure occurring during the execution increases linearly with additional processors. Fault tolerant methods attempt to reduce the expected execution time by allowing recovery from Failure. This paper presents an analysis and implementation of a fault tolerant Cholesky factorization algorithm that does not require checkpointing for recovery from fail-stop Failures. Rather, this algorithm uses redundant data added in an additional set of processes. This differs from previous works with algorithmic methods as it addresses fail-stop Failures rather than fail-continue cases. The proposed fault tolerance scheme is incorpoRated into ScaLAPACK and validated on the supercomputer Kraken. Experimental results demonstRate that this method has decreasing overhead in relation to overall runtime as the matrix size increases, and thus shows promise to reduce the expected runtime for Cholesky factorizations on very large matrices.

  • algorithmic cholesky factorization fault recovery
    International Parallel and Distributed Processing Symposium, 2010
    Co-Authors: Doug Hakkarinen, Zizhong Chen
    Abstract:

    Modeling and analysis of large scale scientific systems often use linear least squares regression, frequently employing Cholesky factorization to solve the resulting set of linear equations. With large matrices, this often will be performed in high performance clusters containing many processors. Assuming a Constant Failure Rate per processor, the probability of a Failure occurring during the execution increases linearly with additional processors. Fault tolerant methods attempt to reduce the expected execution time by allowing recovery from Failure. This paper presents an analysis and implementation of a fault tolerant Cholesky factorization algorithm that does not require checkpointing for recovery from fail-stop Failures. Rather, this algorithm uses redundant data added in an additional set of processors. This differs from previous works with algorithmic methods as it addresses fail-stop Failures rather than fail-continue cases. The implementation and experimentation using ScaLAPACK demonstRates that this method has decreasing overhead in relation to overall runtime as the matrix size increases, and thus shows promise to reduce the expected runtime for Cholesky factorizations on very large matrices.

Dao Zhou - One of the best experts on this subject based on the ideXlab platform.

  • mission profile based system level reliability analysis of dc dc converters for a backup power application
    IEEE Transactions on Power Electronics, 2018
    Co-Authors: Dao Zhou, Huai Wang
    Abstract:

    Reliability analysis is an important tool for assisting the design phase of a power electronic converter to fulfill its life-cycle specifications. Existing converter-level reliability analysis methods have two major limitations: 1) being based on Constant Failure Rate models; and 2) lack of consideration of long-term operation conditions (i.e., mission profile). Although various studies have been presented on power electronic component-level lifetime prediction based on wear-out Failure mechanisms and mission profile, it is still a challenge to apply the same method to the reliability analysis of converters with multiple components. Component lifetime prediction based on associated models provides only a $B_{X}$ lifetime information (i.e., the time when X % items fail), but the time-dependent reliability curve is still not available. In this paper, a converter-level reliability analysis approach is proposed based on time-dependent Failure Rate models and long-term mission profiles. Two different methods to obtain the component-level time-to-Failure are illustRated by a case study of dc/dc converters for a 5 kW fuel cell-based backup power system. The reliability analysis of the converters with and without redundancy is also performed to assist the decision making in the design phase of the fuel cell power conditioning stage.

Stephen P. Harris - One of the best experts on this subject based on the ideXlab platform.

  • The Effects of Maintenance Actions on the Average Probability of Failure on Demand of Spring OpeRated Pressure Relief Valves Contributed by the Pressure Vessel and Piping Division of ASME for publication in the JOURNAL OF PRESSURE VESSEL TECHNOLOGY. Manus
    2020
    Co-Authors: Julia V. Bukowski, William M. Goble, Robert E. Gross, Stephen P. Harris
    Abstract:

    The safety integrity level (SIL) of equipment used in safety instrumented functions is determined by the average probability of Failure on demand (PFDavg) computed at the time of periodic inspection and maintenance, i.e., the time of proof testing. The computation of PFDavg is generally based solely on predictions or estimates of the assumed Constant Failure Rate of the equipment. However, PFDavg is also affected by maintenance actions (or lack thereof) taken by the end user. This paper shows how maintenance actions can affect the PFDavg of spring opeRated pressure relief valves (SOPRV) and how these maintenance actions may be accounted for in the computation of the PFDavg metric. The method provides a means for quantifying the effects of changes in maintenance practices and shows how these changes impact plant safety. In a properly operating SOPRV, a spring exerts a downward force/pressure on the disk pressing the disk against the seat. The seat is the top surface of the wall of the nozzle. The green circles in During normal plant operation, the SOPRV is in the closed position. If the process pressure increases beyond that of the spring set pressure, the disk will be lifted allowing process fluid to flow through the outlet thereby relieving excess process pressure. When the process pressure returns to the closing pressure of the SOPRV, the disk once again closes against the seat to provide a fluid tight seal and the process proceeds normally. The SOPRV can fail in a number of ways. If the SOPRV either spuriously opens or fails to form a fluid tight seal when the process pressure is within normal ranges, the valve is said to leak. These Failure modes usually are considered to be safe Failures (provided that the unintended pressure relief and fluid release do not themselves induce a safety hazard). On the other hand, if the SOPRV does not open under conditions of excessive process pressure, the valve is said to be "fail to open" (FTO) or to be "stuck shut," and this is a dangerous Failure. PFDavg measures the average probability of being in this dangerous Failure mode when excessive process pressure needs to be relieved. In a process, the occurrence of excessive pressure is called a demand on the SOPRV hence the metric, PFDavg. Because the SOPRV is normally closed, it is not possible to observe the FTO dangerous Failure mode during normal operation. Consequently, safety standards, such as Refs. Journal of Pressure Vessel Technology DECEMBER 2015, Vol. 137 / 061601-1 Copyright V C 2015 by ASME identified. Implementation of these maintenance actions will reduce the Failure Rate and impact PFDavg. The remainder of this paper: • provides background information about the computation of PFDavg relevant to the study • describes the source of data, rationale for data choice, and summarizes the relevant data • presents an analysis of the proof test Failure data for a particular group of SOPRV • provides examples of categorizing the FTO and using the results to calculate the necessary parameters for computing PFDavg • shows the impacts of three different levels of maintenance actions on PFDavg under two different assumptions about infant mortality Failures and compares these to the ideal case • closes with a discussion of the results and conclusions

  • The Effects of Maintenance Actions on the PFDavg of Spring OpeRated Pressure Relief Valves
    Volume 6B: Materials and Fabrication, 2014
    Co-Authors: Julia V. Bukowski, William M. Goble, Robert E. Gross, Stephen P. Harris
    Abstract:

    The safety integrity level (SIL) of equipment used in safety instrumented functions is determined by the average probability of Failure on demand (PFDavg) computed at the time of periodic inspection and maintenance, i.e., the time of proof testing. The computation of PFDavg is generally based solely on predictions or estimates of the assumed Constant Failure Rate of the equipment. However, PFDavg is also affected by maintenance actions (or lack thereof) taken by the end user. This paper shows how maintenance actions can affect the PFDavg of spring opeRated pressure relief valves (SOPRV) and how these maintenance actions may be accounted for in the computation of the PFDavg metric. The method provides a means for quantifying the effects of changes in maintenance practices and shows how these changes impact plant safety.