论文部分内容阅读
1973年全国制订《主要农作物病虫预测预报办法》中规定玉米螟的防治标准为:“春玉米心叶中期百株累计卵量超过30块时必须进行防治……。”要作好预测预报、正确指导大田防治工作,必须在玉米地进行查卵。 一、应用序贯分析的必要性 现时测报对卵块调查是以五点取样,每点固定调查20株,工作量较大。如果在卵块少的地块,可无需检查100株,即可凭经验肯定其卵块数不会达到防治指标;同样,在卵块密度很高的地块,其实也无需查到100株就也能大致推断其卵块数能否达到防治标准。在这两种情况下如果都硬性规定调查至100株,便会造成不必要的人力浪费。但调查株数须少到什么程度而我们的判断又能保证不产生错误?可用序贯抽样根据假设检验理论来作出推断。
In 1973 the country to develop “Prediction of major crop pest forecasting method” provides for the prevention and control of corn borer is: “mid-spring corn leaf mid-100 cumulative number of eggs must be controlled ... ...” To make forecasts and forecasts, Correct guidance to the field of prevention and control work, we must check the corn in the corn. First, the application of the need for sequential analysis At present, the survey of eggs on the block is based on five sampling, fixed point survey of 20 points each, a larger workload. If less eggs in the block, you can check without inspection of 100, that can be empirically determined that the number of eggs does not meet the prevention and control targets; Similarly, the high egg mass density plots, in fact, needless to find 100 can be roughly Inferred the number of eggs can reach the prevention and control standards. In both cases, if all are hard-pressed to investigate into 100 strains, unnecessary waste of manpower will be caused. However, the investigation of the number of strains to what extent and our judgments can be guaranteed without errors? Sampling can be used based on the hypothesis to test the theory to make inferences.