在多种无信息先验下,将Gibbs抽样与Metropolis-Hastings算法混合的方法和重要抽样法应用于幂律过程强度函数的Bayesian预测分析,简化Bayesian分析同时还能方便地给出强度函数及其函数的Bayes估计和区间分析.所给预测方法不仅能预测幂律过程的未来强度,同样适用于当前强度的预测。在用具有精确解的数值模拟算例充分验证了文中方法的可行性、合理性和有效性之后,将其应用于一个实例分析,并就无信息先验中参数的选取给出一些建议. Under a variety of non-information priori, the Gibbs sampling and Metropolis-Hastings algorithm and the important sampling method are applied to the Bayesian predictive analysis of the power law process intensity function. Simplifying the Bayesian analysis, the intensity function and its Bayesian estimation of function and interval analysis.The given prediction method can not only predict the future strength of power-law process, but also apply to the prediction of current strength. After the numerical simulation example with exact solution has been used to verify the feasibility, rationality and validity of the proposed method, this method is applied to an example analysis and some suggestions are given for the selection of parameters in a priori informationless.