liveBook · Manning

Online step-by-step pheromone update

As explained in Optimization Algorithms: AI techniques for design, planning, and control problems

An online step-by-step pheromone update deposits pheromone during the incremental construction of a candidate solution.

An online step-by-step pheromone update deposits pheromone on each edge while an ant incrementally constructs a candidate solution. Strengthened trails increase the probability that later ants choose those edges, linking construction with collective search. section 10.2 section 10.6

The update operates within the broader ACO cycle:

  • Evaporate existing pheromone before depositing new trail; evaporation prevents permanent commitment to one route. section 10.2 section 10.6
  • Deposit pheromone immediately on every traversed edge during solution construction. section 10.2
  • Choose the deposit rule: ant density adds a constant Q, while ant quantity makes deposition proportional to local edge quality, such as Q/dij. section 10.2
  • Balance reinforcement against exploration: no evaporation can trap ants in a local optimum, whereas maximal evaporation forces largely random exploration. section 10.2

Routing interpretation

When an ant traverses an edge from node i to node j, the ant quantity model deposits Q/dij. Shorter edges therefore receive stronger reinforcement, while the ant density model rewards edges selected by more ants without considering edge length. section 10.2

Figure 10.5 Pheromone update methods

Compare online step-by-step deposition with delayed pheromone updating after a complete solution has been constructed. section 10.2 section 10.6

How Optimization Algorithms: AI techniques for design, planning, and control problems covers it

How deep it goes
Explains deposit timing, density and quantity rules, and evaporation trade-offs, but defers implementation details beyond these models.
Examples
Uses traversed graph edges, node-to-node distance, and ant food-foraging trails to illustrate reinforcement.
What you can do after
Apply pheromone deposits using constant traffic-based reinforcement or distance-based edge reinforcement while tuning evaporation for exploration.
How it connects
Connects pheromone evaporation and deposit with ant colony optimization, ant density, ant quantity, and delayed pheromone updating.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Optimization Algorithms: AI techniques for design, planning, and control problems. The section links point to the parts of the book it draws on.