@tscircuit/high-density-a01

A high density zero-obstacle solver

This is a @tscircuit/solver-utils BaseSolver-compatible solver with the following properties:

• Multi-layer
• Grid-based
• Supports High Density Types from tscircuit-autorouter
• Rip'n'Replace with History (@tscircuit/hypergraph-inspired)
• Via Penalty and Trace Penalty Map

Usage

const solver = new HighDensitySolverA01({
  nodeWithPortPoints,
  cellSizeMm: 0.05,
  viaDiameter: 0.3,
  stepMultiplier: 4, // optional, each solver step runs 4 internal steps

  // optional hyperparameters to control algorithm
  // Unit of penalty is ~mm
  hyperParameters: {
    shuffleSeed: 0,
    ripCost: 10,
    ripTracePenalty: 0.5,
    ripViaPenalty: 0.75,
    viaBaseCost: 0.1,
  },

  // Optional functions to generate initial penalty map
  // initialPenaltyFn: ({ x, y, px, py, row, col }) => ...
})

solver.solve()

How it works

We form a grid based on the parameters cellSizeMm

We compute the initial penalty map from the initialPenaltyFn, this function sets an additional cost of traversal for a cell. This function accepts x,y which represent the "real coordinates", as well as px/py which are [0,1] representing the fraction of the coordinate within the problem bounds.

We shuffle the trace order based on the shuffle seed.

We run a basic A* solver for each path from the start to the end. When we explore, we consider both used and unused cells via the usedCell structure. The usedCell structure contains 0 or 1 depending on whether or not a cell has been used by either a trace or via. When we explore, we explore in 8 directions as well as creating a via. A via enables exploring to any other layer. When a cell is used, we add the ripCost to the exploration of that path. A path that rips the same trace will only incur the ripCost once, so we must track for each path what traces have been ripped.

When we reach the end of a path, we then mark that route as solved and apply the used cells to the usedCell structure. Note that vias occupy more cells based on the viaDiameter. We then see if we needed to rip any traces. When we rip a trace, we remove it from the usedCell structure, remove it from the solvedConnections list and add it back to the unsolvedConnections queue.