基于LM算法的BP神经网络的电力负荷短期预测

清代书院课艺总集多为连续出版物，或具有连续出版物的刊行初衷。刊期短则一季，多则一年或数年。经费充足与否，会影响刊期。发表周期多为一年至五年，也有十余年的。用稿率以10%~20%居多，偶见“关系稿”。时文的用稿标准是“清真雅正”。题目多为官师所拟。一般全文刊登，也偶有“论点摘编”。多经润色，并附录评点。有的以袖珍本刊行，有的宣称“翻刻必究”，标出定价，附载广告。稿费已在膏火费中预支，优秀作品可被转载。从本质属性和诸多要素来看，书院课艺总集实开今日“大学学报”、“学术集刊”之先河。

Power System Short-Term Load Forecasting Based on Levenberg-Marquardt Algorithm BP Neural Network

Test paper collections of academies in Qing Dynasty are normally serial publication or something like that.According to publication frequency,test paper collections can be classified into quarterly collection,year book and collections published every several years,which largely depends on the publishing funds.The publishing period ranges from one year to five years,sometimes even covers ten years.Publishing rates are mostly 10%-20% and the standards of eight-part essay are purity and elegance.Titles are assigned by officials and teachers.Usually,the entire chapter would be published,while somtimes only one short paragraph could be cited.Most papers are well polished and thoroughly reviewed.Some collections are pocket editions,and some are not allowed to be pirated.Price and advertisements are always indicated.The payment serves as scholarship.Excellent works are available for reprinting.In terms of essential attributes and various elements,test paper collections are the origin of academic journal and collected papers at present.