Mechanical Engineering

Permanent URI for this communityhttp://hdl.handle.net/1903/2263

Browse

Subject: search.filters.subject.algorithm

Search Results

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Algorithm to Determine the Knee Point on Capacity Fade Curves of Lithium-Ion Cells
    (MDPI, 2019-07-29) Diao, Weiping; Saxena, Saurabh; Han, Bongtae; Pecht, Michael
    Lithium-ion batteries typically exhibit a transition to a more rapid capacity fade trend when subjected to extended charge–discharge cycles and storage conditions. The identification of the knee point can be valuable to identify the more severe degradation trend, and to provide guidance when scheduling battery replacements and planning secondary uses of the battery. However, a concise and repeatable determination of a knee point has not been documented. This paper provides a definition of the knee point which can be used as a degradation metric, and develops an algorithm to identify it. The algorithm is implemented on various data cases, and the results indicate that the approach provides repeatable knee point identification.
  • Thumbnail Image
    Item
    Biologically-inspired optimal control
    (2005-11-14) Shao, Cheng; Hristu, Dimitrios; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    Inspired by the collective activities of ant colonies, and by their ability to gradually optimize their foraging trails, this dissertation investigates the cooperative solution of a broad class of trajectory optimization problems with various types of boundary conditions. A set of cooperative control algorithms are presented and proved to converge to an optimal solution by iteratively optimizing an initially feasible trajectory/control pair. The proposed algorithms organize a group of identical control systems by imposing a type of pair-wise interaction known as "local pursuit". The bio-inspired approach taken here requires only short-range, limited interactions between group members, avoids the need for a "global map" of the environment in which the group evolves, and solves an optimal control problem in "small" pieces, in a manner which is made precise. These features enable the application of the proposed algorithms in numerical optimization, leading to an increase of the permitting size of problems that can be solved, as well as a decrease of numerical errors incurred in ill-conditioned problems. The algorithms' effectiveness is illustrated in a series of simulations and laboratory experiments