STRATA
Skip to study
XLVII

X.06 / SHARED INSTRUMENTS

Optimization, control, and decision theory

Which states can a controller observe, and which can it reach within its operating limits?

A source-led introduction1 source references1 related experiments

Start with the question.

Optimization makes an objective and constraints explicit; control adds decisions over time, and decision theory connects those choices to uncertain outcomes and their consequences. Dantzig's linear-programming work offers a concrete mathematical entry point. A useful model makes tradeoffs inspectable instead of hiding them inside a single score, especially when the objective is only a proxy for the intended result.

Solve a small allocation problem.

Let x and y be nonnegative activity levels with x + y ≤ 4, x ≤ 2, and y ≤ 3. Maximize the illustrative value 3x + 2y. The feasible choice x = 2, y = 2 gives value 10; compare the other corner points to verify the optimum. Then change one coefficient and inspect whether the preferred allocation changes.

What would you inspect?

  • Objective coefficients, constraints, feasible region, and sensitivity of the solution.
  • The consequences omitted by the chosen objective or uncertainty model.

Take the next step.

Write an allocation problem from a real research task and test whether its recommended choice survives a plausible change in costs or priorities.

Put part of the question to work.

These models isolate a mechanism you can vary. Their assumptions define the part of this topic they can illuminate.

Try forcing, damping, and response.

Change one parameter, compare the numerical output, and export the result. The experiment loads when you open it.

Open its full page

Watch the quantities change.

12 seconds · Original model visualization · Silent

Forcing, damping, and response

Compare a finite driven motion with the response that persists after transients decay.

Read the explanation and method

A damped oscillator starts at rest and is driven sinusoidally at its natural frequency of 1 hertz. The damping ratio is 0.12.

Cream shows normalized displacement and orange shows the normalized applied force. The transient grows toward a periodic response.

The analytical steady-state gain is about 4.17. The displayed finite trajectory comes from fourth-order Runge-Kutta integration and may retain a transient.

Method. Integrate the normalized mass–spring equation with fourth-order Runge–Kutta steps of 1/240 second. The plotted trace keeps 60 samples per second. Separately calculate the analytical steady-state gain and phase for a continuously driven linear system.

Read the supporting material.

  1. Maximization of a Linear Function of Variables Subject to Linear Inequalities

    George B. Dantzig / Cowles Commission · Linear programming and the simplex framework in an original chapter.

Continue through the versioned study records.

Follow the comparisons.

Read the neighboring questions.