Aiming at the shortcomings of the traditional reliability evaluation of electronic products, a life prediction method of electronic products was proposed which was based on the failure of physics and pseudo-failure data. First of all, based on the failure mode, failure mechanism analysis of electronic products, the sensitive performance parameters of the product was determined; then by monitoring the degradation amount of sensitive parameters, the model of electronics degradation path was built, and maximum likelihood method was used to estimate its parameters; finally, by setting the fault threshold value, the life distribution of electronic products was gained. Simulation showed that the evaluation accuracy of the method is higher and providing a new way to evaluate the reliability of electronic products.

Pointing at the problem that the spares consumption quota has been using the experience to develop, which makes spares application random and blind, put forward to build life-repairable spares consumption quota model, to reasonably develop consumption quota of this kind of spares. Analyze and determine the factors influencing the life-repairable spares consumption, use BP neural network to predict, and use genetic algorithm to optimize the weights and thresholds of BP neural network, so that the network can obtain the global minimum point. The example shows that the model's predicted results are relatively accurate and has high practicability.