All Superinterfaces:
OptimizationData

public interface SQPProblem extends OptimizationData
Defines a nonlinear constrained optimization problem suitable for use with Sequential Quadratic Programming (SQP) solvers.

Implementations supply objective function information, constraint definitions, bounds, and problem dimensionality. SQP solvers use this interface to query the optimization model.

Each component is optional: the problem may include only an objective (unconstrained case), or objective plus bounds, or full equality and inequality constraint sets. Boolean guard methods allow users and solvers to detect which problem structures are active.

Implementations must be deterministic and thread-safe if used across multiple optimizer instances.

Since:
1.0
  • Method Details

    • getdim

      int getdim()
      Returns the number of decision variables of the problem.
      Returns:
      problem dimension
    • hasBounds

      boolean hasBounds()
      Indicates whether explicit variable bounds are defined.
      Returns:
      true if box constraints are available
    • getBoundsLB

      double[] getBoundsLB()
      Returns lower bounds for each decision variable, if available.
      Returns:
      array of lower bound values (length equals problem dimension)
    • getBoundsUB

      double[] getBoundsUB()
      Returns upper bounds for each decision variable, if available.
      Returns:
      array of upper bound values (length equals problem dimension)
    • getObjectiveEvaluation

      double getObjectiveEvaluation(RealVector rv)
      Evaluates the objective function at the given point.
      Parameters:
      rv - point at which to evaluate the objective
      Returns:
      objective function value
    • getObjectiveGradient

      RealVector getObjectiveGradient(RealVector rv)
      Evaluates the gradient of the objective function at the given point.
      Parameters:
      rv - point at which to evaluate the gradient
      Returns:
      gradient vector
    • hasEquality

      boolean hasEquality()
      Indicates whether equality constraints are present.
      Returns:
      true if the problem contains equality constraints
    • getEqConstraintEvaluation

      RealVector getEqConstraintEvaluation(RealVector rv)
      Computes equality constraint values at the given point.
      Parameters:
      rv - evaluation point
      Returns:
      vector of equality constraint residuals
    • getEqConstraintJacobian

      RealMatrix getEqConstraintJacobian(RealVector rv)
      Computes the Jacobian of the equality constraints at the given point.
      Parameters:
      rv - evaluation point
      Returns:
      Jacobian matrix of equality constraints
    • getEqConstraintLB

      RealVector getEqConstraintLB()
      Returns the lower bound for equality constraints (typically zero).

      Some interior-point/SQP formulations treat equalities as bounded two–sided constraints h(x) = 0.

      Returns:
      vector of equality constraint lower bounds
    • hasInequality

      boolean hasInequality()
      Indicates whether inequality constraints are present.
      Returns:
      true if the problem contains inequality constraints
    • getIneqConstraintEvaluation

      RealVector getIneqConstraintEvaluation(RealVector rv)
      Computes inequality constraint values at the given point.

      Inequalities follow the canonical form g(x) >= c.

      Parameters:
      rv - evaluation point
      Returns:
      vector of inequality constraint residuals
    • getIneqConstraintJacobian

      RealMatrix getIneqConstraintJacobian(RealVector rv)
      Computes the Jacobian of inequality constraints at the given point.
      Parameters:
      rv - evaluation point
      Returns:
      Jacobian matrix of inequality constraints
    • getIneqConstraintLB

      RealVector getIneqConstraintLB()
      Returns the lower bound for inequality constraints.
      Returns:
      vector of lower bound values for inequality constraints