CAJ | 학술논문

为解决农业试验统计分析软件、扬州分析器（2.2）、Excel进行回归分析建立的部分非典型性肥料效应方程推荐的施肥量出现异常偏大、偏低甚至负值，脱离当地生产实际的问题。该文应用概率理论的基本知识和频数的原理和方法，对油菜田间肥效试验数据应用频率分析法进行统计分析，得到的油菜优化施肥量产量为1732.4 kg/hm2，氮、磷、钾优化组合合范围分别为210.36~149.64、81.89~58.11、81.89~58.11 kg/hm2，与试验地（田）和当地生产实际相吻合。该研究是在频率分析的基础上，用加权平均方法，确定各种不同产量目标的生产因素组合，将其中产量高，出现频率大（稳产），用肥节省的组合作为优化生产措施，具有增加施肥决策信息量，减少或避免小概率事件的风险的优点。可解决肥效试验数据分析统计汇总应用扬州分析器（2.2）、Excel、DPS等统计分析软件进行回归分析建立的部分非典型肥料效应函数推荐的施肥量和目标产量出现异常值，与当地生产实际不相符的问题。肥料效应函数方程经方差分析不显著的试验，能否应用该研究方法进行施肥决策，推荐氮、磷、钾肥料优化组合和优化施肥量产量，还需在以后的工作实践中进一步探讨。
위해결농업시험통계분석연건、양주분석기（2.2）、Excel진행회귀분석건립적부분비전형성비료효응방정추천적시비량출현이상편대、편저심지부치，탈리당지생산실제적문제。해문응용개솔이론적기본지식화빈수적원리화방법，대유채전간비효시험수거응용빈솔분석법진행통계분석，득도적유채우화시비양산량위1732.4 kg/hm2，담、린、갑우화조합합범위분별위210.36~149.64、81.89~58.11、81.89~58.11 kg/hm2，여시험지（전）화당지생산실제상문합。해연구시재빈솔분석적기출상，용가권평균방법，학정각충불동산량목표적생산인소조합，장기중산량고，출현빈솔대（은산），용비절성적조합작위우화생산조시，구유증가시비결책신식량，감소혹피면소개솔사건적풍험적우점。가해결비효시험수거분석통계회총응용양주분석기（2.2）、Excel、DPS등통계분석연건진행회귀분석건립적부분비전형비료효응함수추천적시비량화목표산량출현이상치，여당지생산실제불상부적문제。비료효응함수방정경방차분석불현저적시험，능부응용해연구방법진행시비결책，추천담、린、갑비료우화조합화우화시비양산량，환수재이후적공작실천중진일보탐토。
To solve the problems of abnormal larger, abnormal lower or even nega-tive of target yield and fertilizing amount recommended by part of non-typical fertil-izer effect equations using agricultural experiments and statistical analysis software, Yangzhou analyzer (2.2), regression analysis of Excel, which objected to local actual production, the study adopted the principle and method of basic knowledge and the frequency of using probability theory, and carried out statistical analysis on the rape field fertilizer experiment data by frequency analysis method, the rape yield after op-timizing fertilizing amount was 1 732.4 kg/hm2, the ranges of N, P and K optimal combinations were: N=210.36-149.64 kg/hm2,P2O5=81.89-58.11 kg/hm2, K2O=81.89-58.11 kg/hm2,which was consistent with local actual production. This study was based on frequency analysis, using weighted average method to determine the pro-duction combinations of different yield objectives, hereinto, the combinations with high yield, high frequency of occurrence (dependable crop) and fertilizer-saving were viewed as the optimizing production measures, and they had the merits of increas-ing fertilization decision-making information, reducing or avoiding the risk of smal probability event. The results of this study can solve the problem of abnormal val-ues fertilizing amount and target yield recommended by non-typical fertilizer effect function, which did not accord with local actual production, caused by Yangzhou an-alyzer (2.2), regression analysis of Excel, and DPS statistical analysis software. For the fertilizer effect function equation established by regression analysis which did not reach significance level using variance analysis, whether the method can be adapt-ed to for carrying out fertilization decision-making, recommending optimization com-binations of N, P and K fertilizers and yield under optimized fertilizing amount should be further researched in future working practice.