Applications of Pseudorandom Number Generator in Selected Stochastic Simulations

The accuracy of percolation models depends heavily on the quality of the pseudorandom number generator (PRNG) used. Traditional PRNGs may show periodicity, correlations, or biases, affecting simulation reliability. We propose a novel PRNG based on chaotic maps—the A2 attractor and the Ikeda map—to improve randomness. Its performance, evaluated with statistical test suites, exceeds conventional standards. Applied to site percolation on a square lattice, the PRNG provides reliable outputs, suitable for Monte Carlo methods and complex systems modeling.