Most of traditional reliability models for traditional mechanical products ignore the oscillation characteristics of probability density caused by periodic faults， which affects the accuracy of the models. The reliability model of the exponential attenuation oscillation distribution and the corresponding reliability， failure rate， and average life calculation models are constructed on the basis of the traditional life model. Then， the effect of parameters including the attenuation coefficient， oscillation amplitude， oscillation angular frequency on the probability distribution characteristics is studied. As an example， the failure time data of automatic transmission of a forklift truck are fitted with the proposed probability distribution model. And the fitting results are compared with those of the exponential distribution model， the three-parameter Weibull model and the bathtub-shaped curve model. The results show that the probability model can better describe the probability distribution characteristics of the fault-free working time of automatic transmission， and the model can obtain a smaller root mean square error.