2019/09/11 by Tijana Janjić, Yvonne Ruckstuhl, Janjic, Tijana +3
Computer Science · Mathematics · #65K05 #86A10 #90C20 #Advanced Multi-Objective Optimization Algorithms #Advanced Optimization Algorithms Research #FOS: Mathematics #FOS: Physical sciences #G.1.6 #J.1 #Mathematical Physics (math-ph) #Matrix Theory and Algorithms #Optimization and Control (math.OC) #Solar Radiation and Photovoltaics
paper · pdf · doi:10.48550/arxiv.1909.04991
openalex publication_date 2019/09/11 · openalex created_date 2022/07/28 · openalex updated_date 2026/07/28
A specialized algorithm for quadratic optimization (QO, or, formerly, QP)\nwith disjoint linear constraints is presented. In the considered class of\nproblems, a subset of variables are subject to linear equality constraints,\nwhile variables in a different subset are constrained to remain in a convex\nset. The proposed algorithm exploits the structure by combining steps in the\nnullspace of the equality constraint's matrix with projections onto the convex\nset. The algorithm is motivated by application in weather forecasting.\nNumerical results on a simple model designed for predicting rain show that the\nalgorithm is an improvement on current practice and that it reduces the\ncomputational burden compared to a more general interior point QO method. In\nparticular, if constraints are disjoint and the rank of the set of linear\nequality constraints is small, further reduction in computational costs can be\nachieved, making it possible to apply this algorithm in high dimensional\nweather forecasting problems.\n