Devil’s Trumpet Optimization (DTO)

This is new bio-inspired optimization algorithm based on the Datura plant

Biological Inspiration

1. Night Blooming & Pollination

• Datura flowers open at night, attracting moths.
• This can inspire exploration (global search) through “random pollination flights.”

1. Selective Pollination

• Moths prefer nectar-rich flowers (high fitness).
• This represents local exploitation (intensifying search near best solutions).

1. Seed Dispersal

• Thorn apple fruits disperse many seeds randomly.
• This introduces mutation/diversification in candidate solutions.

1. Toxicity Defense

• Datura is poisonous; not every interaction is beneficial.
• This can be modeled as a penalty mechanism to reject poor or infeasible solutions.

🧮 Algorithm Steps (DTO)

1. Initialization

• Randomly generate a population of flowers (candidate solutions).

1. Global Pollination (Exploration)

• Use Lévy flight or random walk to generate new positions.

Xit+1=Xit+β⋅L(λ)⋅(Xbest−Xit)X_i^{t+1} = X_i^t + \beta \cdot L(\lambda) \cdot (X_{best} - X_i^t)Xit+1​=Xit​+β⋅L(λ)⋅(Xbest​−Xit​)

1. Local Pollination (Exploitation)

• Select two flowers randomly; exchange information:

Xit+1=Xit+ϵ(Xjt−Xkt)X_i^{t+1} = X_i^t + \epsilon (X_j^t - X_k^t)Xit+1​=Xit​+ϵ(Xjt​−Xkt​)

where ϵ∼U(0,1)\epsilon \sim U(0,1)ϵ∼U(0,1).

1. Seed Burst (Mutation)

• Occasionally, a solution “explodes” into multiple variants (like thorn apple seeds).
• Introduce diversity to escape local minima.

1. Toxicity Screening (Constraint Handling)

• Apply penalties for infeasible/low-fitness candidates.

1. Selection & Elitism

• Keep the best solutions and repeat until stopping criteria.