Methods
Algorithms, heuristics, and geometry primitives used to solve problems.
- Phi-functions
Analytic non-overlap modeling that encodes the interaction of two pieces.
- First-Fit Decreasing
Sort items largest-first, then place each in the first fitting bin.
- Reinforcement Learning
Learning placement/sequencing policies from reward signals.
- Inner-Fit Polygon
Geometric primitive describing where a piece fits inside the stock boundary.
- Dynamic Programming
Exact technique that stores overlapping subproblems in a table.
- No-Fit Polygon
Geometric primitive describing all overlap-free relative placements of two pieces.
- Minkowski Sum
Geometric operation underlying No-Fit Polygon computation.
- Bottom-Left Fill
Constructive placement heuristic that drops each piece to the lowest-leftmost feasible position.
- Branch and Bound
Exact tree search that prunes subproblems using bounds.
- Simulated Annealing
Single-solution metaheuristic that accepts worse moves with decreasing probability.
- Column Generation
Solve huge LPs by generating variables (columns) on demand.
- Genetic Algorithm
Population-based metaheuristic searching over encoded solutions.
- Integer Linear Programming
Formulating cutting & packing as a linear model over integer variables.
- Tabu Search
Local search with a memory of recently visited moves to avoid cycling.
- Lower Bounds
Bounding the optimum to prune exact search and certify solution quality.