By Alexander S. Belenky (auth.)
The medical monograph of a survey style provided to the reader's cognizance bargains with basic principles and simple schemes of optimization tools that may be successfully used for fixing strategic making plans and operations deal with ment difficulties comparable, specifically, to transportation. This monograph is an English translation of a substantial a part of the author's e-book with an identical name that used to be released in Russian in 1992. the fabric of the monograph embraces equipment of linear and nonlinear programming; nonsmooth and nonconvex optimization; integer programming, fixing difficulties on graphs, and fixing issues of combined variables; rout ing, scheduling, fixing community circulation difficulties, and fixing the transportation challenge; stochastic programming, multicriteria optimization, online game idea, and optimization on fuzzy units and lower than fuzzy pursuits; optimum keep an eye on of structures defined by way of traditional differential equations, partial differential equations, gen eralized differential equations (differential inclusions), and sensible equations with a variable which could think in simple terms discrete values; and a few different equipment which are according to or adjoin to the indexed ones.
Read or Download Operations Research in Transportation Systems: Ideas and Schemes of Optimization Methods for Strategic Planning and Operations Management PDF
Best linear programming books
Linear matrix inequalities (LMIs) have lately emerged as necessary instruments for fixing a couple of keep an eye on difficulties. This e-book offers an up to date account of the LMI strategy and covers themes akin to contemporary LMI algorithms, research and synthesis concerns, nonconvex difficulties, and functions. It additionally emphasizes functions of the tactic to components except regulate.
Integer recommendations for structures of linear inequalities, equations, and congruences are thought of in addition to the development and theoretical research of integer programming algorithms. The complexity of algorithms is analyzed based upon parameters: the measurement, and the maximal modulus of the coefficients describing the stipulations of the matter.
- Optimization with Multivalued Mappings: Theory, Applications and Algorithms
- Iterative methods for linear and nonlinear equations
- Linear and nonlinear programming
- The Geometry of Higher-Order Lagrange Spaces: Applications to Mechanics and Physics
- Linear Programming and Algorithms for Communication Networks: A Practical Guide to Network Design, Control, and Management
- An invitation to variational methods in differential equations
Extra info for Operations Research in Transportation Systems: Ideas and Schemes of Optimization Methods for Strategic Planning and Operations Management
Iterative methods of linear programming have several advantages over finite methods of the simplex type. First, the initial information concerning the problem can be written in a more compact form (it reduces the number of operations in working with matrices of the constraints that contain many nonzero elements). Second, these methods do not require inverting the matrix (it cuts down the operating memory and hence allows one to solve problems of larger dimensions). Third, a solution can be rapidly corrected in altering the initial 24 CH.
The new vertex is chosen out of the points Xk and 2x~ - x~g at the point at which f(x) assumes the smaller value, and if then x~ is taken as the new vertex. If the value of f(x) at the point x~ is greater than that at the other simplex vertices, a step towards x~g is made. , is not large), and the initial function f(x) is approximated within it. Otherwise, x~ is taken as the new simplex vertex . One-dimensional minimization methods are also considered as direct search methods. In conformity to unimodal functions considered on a segment (as it was mentioned above), these methods can construct a sequence of segments containing each other and contracting to the minimum of f(x) on the segment.
Some one-dimensional minimization methods are also zero-order ones, the most popular being the Fibonacci and golden section methods. The idea underlying many one-dimensional methods consists of constructing a sequence of intervals imbedded in each other (nested intervals) and tightening to a point of minimum of the goal function. However, such a sequence can be usually constructed for only certain classes of functions. , continuous functions having a unique minimum point in the interval. 3. 3. The most widely used first-order methods include: methods of feasible directions, the conditional gradient method, the gradient projection method, the linearization method, the cutting-plane (or cutting-hyperplane) method, and the penalty function methods.