| Nowadays, the exploitation of clean power and renewable energy resources has become a comprehensive and strategic issue in our country which is involved with the fields of economy, the environment, and energy. However, the implementation of the strategy and the development of the scale are constrained by the low reliability of new energy power systems. Due to the volatility of the wind, solar, tide themselves, power fluctuation produces heat that makes a big impact on the power modules continuously, which expedites the failing process and reduces its reliability, one of the key factors of new energy power generation systems. That to find and build a model which can predict power module residual life(residual life of power module) accurately is the foundation of stable and sustained operation in new energy power generation system. Presently few relative research is carried out on the recession and reliability which caused by thermal shock. There are still many problems in failure process and failure modes cannot be determined clearly. Most of the research is confined to early-warning and cannot predict the practical situation quite accurately.Therefore, much research has been going on in this paper around the on-line prediction of IGBT power modules which are widely used in photovoltaic and wind generation. The main work is as follows:(1) The major factors of IGBT’s failures have been studied and summarized. The three major factors affecting reliable operation are shown below: inner structure defects and manufacturing processes, external stress as well as other environmental conditions. Built upon this basic, the gradually changes between different internal materials and different structures has been accurately elaborated, which are caused by thermal shock.(2)This paper raises and constructs a life prediction model of power IGBT module which is based on real-time temperature monitoring. Through deeply analysis of the existing methods, such as RC thermal network, finite element, and thermal electrical parameters, etc., the trouble is that they can warn of the system, but cannot judge whether the module is really damaged, let alone predicting the remaining life of the module. Aiming at this problem, this paper attempts to propose and construct a model based on real-time monitoring temperature of power IGBT module, which can quantify the total life of the module and immediate consumption module life, and monitor the real-time life changes through cumulative damage during the operation of the module to predict the residual life of power IGBT module.(3) A closed-loop and detecting-temperature system has been set up to simulate practical work and provide really reliable data for the prediction model of power module. To use the cloud model to solve the randomness and fuzziness of IGBT power module’s life, we start with the practical situation and setup a closed-loop and detecting-temperature system with CM100DY-24 NF to carry out series of reliability of aging tests. Some temperature dynamic wave patterns are simulated experimentally to explore the causes of power modules at different temperatures. The cloud model is proposed and introduced in reliability analysis, which, compared with the conventional reliability method, is an effective tool in transforming between qualitative concepts and their quantitative expressions. The random and fuzzy quality in the parameters is showed quantitatively which enhances rationality theoretically.This paper analyses the failure mechanism of IGBT power module in detail and attempts to propose and construct a model based on real-time temperature monitoring of power IGBT module. It is a foundation for the on-line prediction of key power modules in new energy power generation systems. |
PDF Full Text Request