Gradient projection

The constrained minimization problem stated above may be transformed into a form well-suited to gradient projection methods of nonlinear programming by making the following substitution ... 

Accountability of the variances and covariances of the responses makes this optimization procedure particularly noteworthy. From the formulator s viewpoint, the distance criterion could lead to an unacceptable optimum if the formulation levels at the optimum produce response values in an undesirable property range. Khuri and Conlon mentioned the possible use of this procedure for multiresponse mixture optimization although no elaboration or examples were given. The reliance of this method on the gradient projection technique could present difficulties with component level compensation if applied to formulations. 

The above subproblem can be solved very efficiently for fixed values of the multipliers X and v and penalty parameter p. Here a gradient projection trust region method is applied. Once subproblem (3-104) is solved, the multipliers and penalty parameter are updated in an outer loop and the cycle repeats until the KKT conditions for (3-85) are satisfied. LANCELOT works best when exact second derivatives are available. This promotes a fast convergence rate in solving each... 

Rosen, J. B. (1961). The gradient projection method for non-linear programming, Part II, Non-linear constraints. J. Soc. Indus. Appl. Math., 9, 414-32. 

From the theoretical viewpoint, the scalar value of the partial derivate d /dpj is the vector gradient projection to the axis p . Indeed, it can be described with their module and the spatial angle between the vector of the gradient and the axis pj as in relation (3.215) ... 

If the search direction is the gradient direction only, the method joins the family of methods known as the Gradient Projection Methods. 

Methods that use the direction of the function gradient as the projection vector on the active constraints are known as Gradient Projection Methods. 

Note that the minimum found along the gradient projection is very far from and worse than the one obtained with the KKT direction. Since both the points are feasible, the one achieved with the KKT direction is the solution. The point achieved with the gradient projection lies on the bottom of a relatively narrow valley and would struggle to find the solution if only the gradient directions were adopted. 

Quadratic programming is discussed in Chapter 11. The quadratic programming methods are implemented by handling the numerical evaluation of Hessian in a novel way. Reduced gradient and gradient projection are described as conventional methods they are then compared to the novel proposed approach based on the Karush-Kuhn-Tucker direction projection method. It represents the basic core for the development of successive quadratic programming (SQP) methods. 

UK Gradient projects Computer science Internet documents http //gradients.lboro.ac. uk/projects/Batch/index.php. 

Several optimization algorithms, such as the feasible direction method, the gradient projection method, and the linearization method are coded in a multi-purpose program ADS (advanced Design Synthesis) [16], which was used in the present work. A modified feasible directions method [17], was selected in the present numerical example. 

---