Exactness and algorithm of an objective penalty function2012-04-040 PRE AI DOI:10.1007/s10898-012-9900-9原文链接原文求助分享收藏摘要 En