The Methods And Applied Technology Research Of Data Fusion

Much measure will be used to examine the same object in actual engineering test, thenthe data be fused, so that we can make a better estimation of the quality of the object.For a long time, people often provide the point estimator of the measurement whenfusing the data, i.e. they only estimate the mean-value of the measurement. But to judgewhether an engineering is eligible or not,the estimator of mean-value is insufficient,besidesthis, we also need the confidence limit and confidence interval of mean-value. This textarrive at following expression using data fusion method:(1) The expression to estimate mean-value of normal distribution:which wasdeduced with knowing variance(large sample) and unknowing variance(small sample), andit was testified unbias by the theory.(2) The expression to estimate confidence limit of normal distribution mean-value:which was deduced with knowing variance(large sample) and unknowing variance(smallsample).(3) The expression to estimate confidence interval of normal distribution mean-value:which was deduced with knowing variance(large sample) and unknowing variance(smallsample).At the same time, to two variables,with the given degree of confidence, we havededuced the condition of sample variance, which can ensure the new method to have ahigher precison when to estimate the confidence interval of mean-value.When deducing the expression of unknowing variance, it involves dealing with sumof distribution variables. This text use JC method to deal with the question. It avoidedtalking about the sum of t distributed variables directly also can resolve the needingexpression.