Study on Acceptance Criterion of Concrete Strength Based on Bayesian Method
Based on the theory of one-time sampling inspection,balancing the risks of the manufacturer and the user,and considering the protection of the interests of the manufacturer,it is proposed that the evaluation coefficient k for strength average value should be corresponded to 90%the equivalent assurance rate(0.9 Quantile value)for both the known standard deviation method and the unknown standard deviation method,and the sampling scheme when the standard deviation is unknown is determined based on Bayesian theory.When the sampling size is very small,the acceptance criteria of non-statistical method is established based on the average quality standard deviation,and the minimum value acceptance criterion is proposed according to the theory of one-time sampling inspection.Using the concept of Bayesian method,the samples in the previous inspection cycle and the current inspection cycle are combined to expand the sample size and improve the test efficacy of the acceptance criteria with known standard deviation.Bayesian method is adopted to comprehensively consider the average quality standard deviation and the standard deviation of measured samples to improve the test efficacy of acceptance criteria of non-statistical method.The acceptance criteria proposed in this paper have a more clearly statistical basis.The unified equivalent guarantee rate is used to keep the test efficacy close to each other,and the Bayesian theory is used to unify the various methods into the statistical method with unknown standard deviation.The case analysis shows that the acceptance criteria proposed in this paper are slightly stricter than those in GB/T 50107-2010,but the variation range is within the acceptable range;after adopting reasonable prior standard deviation,the Bayesian improved method proposed in this paper can obtain scientific and reasonable evaluation results.
万方数据
维普
