Constrained optimization Created 2026-09-24 Updated 2026-11-05
Constrained optimization minimizes or maximizes an objective function over a feasible set defined by equality or inequality constraints. In subject to and , a feasible optimum must balance improvement of the objective against the permitted directions of motion.
For example, minimizing subject to gives , since . The unrestricted minimizer is infeasible. Lagrange multipliers and the Karush-Kuhn-Tucker conditions generalize this balancing argument; necessity requires appropriate hypotheses, and sufficiency needs extra structure such as convexity.[ref]
Convex Optimization by Stephen Boyd and Lieven Vandenberghe. Chapters 4–5, optimization problems and optimality conditions
Elizabeth Wong (mathematician) Created 2026-10-31 Updated 2026-11-05
Elizabeth Wong coauthored Sequential Quadratic Programming Methods by Philip E. Gill and Elizabeth Wong with Philip E. Gill. The review concerns constrained optimization, in which successive local quadratic problems determine steps and multiplier estimates.
Their work connects the design of optimization algorithms with the linear algebra needed to solve the local subproblems. The 2010 report's title page identifies the authors and institutional report number.[ref]
Lift-constrained maximum lift-to-drag ratio 2026-11-05
A lift-constrained maximum lift-to-drag ratio is the best ratio available within an allowed range of lift coefficients, rather than over every coefficient admitted algebraically by a drag polar. It is a constrained optimization problem whose feasible set must be specified before claiming an attainable optimum.
For the parabolic drag polar with positive and , defineIts derivative isThe ratio increases up to and decreases afterward. The exact constrained solution is thereforeThis proof handles an interior optimum and either active endpoint; simply inserting an upper limit into the unconstrained formula does not.
An original example takes and , giving and unconstrained ratio 15.8114. If the allowed interval is , the maximum occurs at 0.45 and is about 14.9378. If the allowed interval is instead , it occurs at the lower endpoint 0.90 and is about 14.8760. A lower bound can be active as well as an upper bound.
In steady level flight, a speed interval can induce a lift-coefficient interval. With fixed gravitational weight , mass density and wing area ,The endpoints reverse because the function decreases with speed. Additional limits, including the usable range before aerodynamic stall, require intersecting feasible intervals.
The simple finite-wing contribution motivates a quadratic polar, but changing configuration, Mach number or Reynolds number can change its fitted parameters.[ref] Restricting the mathematical interval does not validate the polar across that interval. Wind and ground-distance objectives also define a different optimization from the aerodynamic ratio considered here.
Mathematical optimization Created 2026-09-24 Updated 2026-10-13
Mathematical optimization turns a preference into a precise question: among the allowed decisions, which gives the best outcome? A model separates the objective function, which measures quality, from the optimization constraints, which determine the feasible set. This language unifies finding a cheap plan for a transportation problem, efficient scheduling, or an accurate model in linear regression. The separation is useful because changing a preference and changing what is possible are different modeling decisions. See Boyd and Vandenberghe, introductory optimization models.
For an objective function , a maximization problem asks for a feasible point satisfying for every ; a minimization problem reverses the inequality. This is a global extremum. A local maximum or local minimum compares only nearby feasible points, so finding one does not settle the global question.
Consider allocating ten units between two complementary inputs:The budget constraint forces a tradeoff: an extra unit of costs a unit of . The best allocation is , with objective function value . Without the budget constraint, gives : there is no global maximum. Constrained optimization can therefore have a finite answer even when the corresponding problem with fewer optimization constraints does not. Its beauty is that geometry, calculus, and an arithmetic-geometric mean inequality can certify the same answer.
Maximum and minimum 2026-10-13
Maxima and minima are extrema of a real-valued function. A global maximum of on its domain is a point with for every ; a global minimum reverses the inequality. A local extremum requires the inequality only in a neighbourhood of the point within . In constrained optimization, is the feasible set, so an extremum is tested relative to the constraints rather than in every ambient direction. Thus a constrained local maximum or local minimum can occur at a boundary point without being an interior critical point.
Consider a cylinder of radius , height , and a hemispherical cap of the same radius, enclosing fixed volume . Let the cylindrical wall cost per area and the cap cost per area, with no floor cost. Then and . Eliminating gives , on .
For , its interior minimum satisfies the displayed formulas; the positive second derivative proves uniqueness. For , cost decreases throughout the feasible interval up to its endpoint, so the minimum has and is a hemisphere alone. At cap-to-wall cost ratio two, the optimal cylinder height is its diameter, . This boundary check is part of the constrained optimization, since a formal critical point can otherwise require negative height.
Optimization constraint Created 2026-10-06 Updated 2026-10-20
An optimization constraint restricts the feasible points in constrained optimization, for example by an equality or an inequality . Extrema are considered on the resulting feasible set rather than on the whole domain of the objective.
Past exam of the mathematics course of the University of Cambridge 2003 ib Paper 4 5H Solution Created 2026-09-24 Updated 2026-10-14
For constrained optimization in minimization form, let , and , and define the LagrangianThe Lagrangian sufficiency theorem says: if is feasible, , for every , and minimizes globally over , then minimizes over the feasible set. Equality multipliers may have either sign.
Indeed, for every feasible ,The first inequality uses the inequality-multiplier signs and the equality constraints; the last equality uses complementary slackness. This proves global optimality without assuming convexity or differentiability. The maximization version reverses the relevant signs, or follows by replacing with .
When is convex and the Lagrangian is differentiable and convex, the first-order condition implies its global minimum, because . Thus feasible stationarity plus complementary slackness is sufficient in that convex setting. Mere stationarity in a general nonconvex problem is not the theorem's global-minimum hypothesis.
Past exam of the mathematics course of the University of Cambridge 2019 iii Paper 344 3 a Solution Created 2026-10-03 Updated 2026-10-05
Use the stable parameters . The bulk stationary points satisfy . For , has negative curvature, while the two minima areBoth have and . To establish minimization at fixed total composition, writeFor mean composition with , a mixture of the two minima with volume fraction in the positive phase attains this lower bound in the bulk. Macroscopic slabs have only a subextensive interfacial excess free energy, so the bound is attained per unit volume in the thermodynamic limit. Thus constrained minimization gives phase coexistence in the mean-field approximation; at the endpoints only one phase is needed.
The two equilibrium conditions also follow directly by constrained optimization. Introduce a Lagrange multiplier for in the mixture's bulk free-energy density. Varying and gives ; varying gives . These equate the chemical potentials and pressures, respectively.
Equivalently, their bulk chemical potentials are both zero and their pressures are equal:These are precisely the conditions of the common-tangent construction for phase coexistence: the horizontal supporting tangent touches both minima. Equal chemical potentials prevent net composition exchange, and equal pressure gives mechanical equilibrium at a flat interface.
A central cross-section of the sphere gives the constraint , and the cylinder's volume is . At an interior optimum with , the Lagrange multiplier equations for this constrained optimization problem areThe first gives ; substitution into the second gives . The constraint then givesThe feasible nonnegative form a compact arc, whose endpoints have zero volume. Since this is the only positive stationary candidate, it is the global maximum. Its volume and fraction of the sphere's volume are
Second-derivative test Created 2026-09-28 Updated 2026-10-14
Let have continuous second partial derivatives near an interior critical point of an open subset of , with . The Hessian matrix gives the following sufficient conditions:
- All eigenvalues positive: a strict local minimum.
- All eigenvalues negative: a strict local maximum.
- Both positive and negative eigenvalues: a saddle point of a scalar function.
A positive semidefinite or negative semidefinite Hessian matrix with a zero eigenvalue does not settle the classification. The higher-order test for separated extrema gives two quartic examples with exactly the same zero Hessian. These conditions apply to an interior critical point; a boundary optimum or a constrained optimization problem requires attention to the allowed directions. [ref]
https://ocw.mit.edu/ans7870/18/18.013a/textbook/chapter11/section03.html. MIT 18.013A Calculus with Applications, §11.3: Hessian eigenvalues and the critical-point criterion
The proof outline is the Taylor expansionFor a positive-definite matrix, the smallest eigenvalue supplies a positive bound proportional to , which dominates the remainder for small enough . Negate the function for the local maximum case. For opposite-sign eigenvalues, choose the two corresponding eigenvectors: the leading terms have opposite signs along their lines.
There are also necessary conditions. At an interior local minimum, the Hessian matrix is positive semidefinite; at an interior local maximum, it is negative semidefinite. Restriction to each straight line reduces this to the one-variable second derivative condition. Since the Laplacian is the trace of the Hessian, it is nonnegative at such a local minimum and nonpositive at such a local maximum. The sign of the Laplacian alone is insufficient: has negative Laplacian but a saddle point of a scalar function at the origin.