摘要
均匀实验设计是对微波效应研究现行的全面实验方法的改进,从实验参数空间中挑选有代表性的实验点进行实验,以减少实验所需样本量,降低实验成本。对低噪声放大器的微波效应实验进行两因素六水平的均匀设计,与同条件下的全面实验方法相比理论上节省75%的实验样本量。经仿真实验证明,采用均匀设计结合实验后数据处理,可达到与全面实验类似的效果,预测平均相对误差小于5%。
Uniform design is a method choosing parts of cross-sectional experiment from parameter space.Comparing with actual comprehensive test method,this improved method can decrease the sample size and cut the cost of test.Uniform design was used for making the test plan of microwave effects of Low Noise Amplifier(LNA),which contained two parameters and six values each.The sample size of uniform design was largely decreased by 75% compared to comprehensive test theoretically under the same condition.Simulation results indicate that,the combination of uniform design and succedent data processing can reach similar effect with comprehensive test,and the average comparative error of prediction is below 5%.
出处
《信息与电子工程》
2012年第5期555-558,共4页
information and electronic engineering
基金
国家"863"高技术发展计划资助项目
关键词
均匀设计
微波效应实验
小样本
低噪声放大器
uniform design
microwave effects experiments
small sample size
Low Noise Amplifier
