An Improved Crested Porcupine Optimizer for Path Planning of Mobile Robot

摘要

To address the problem of easily falling into local optimization and low convergence accuracy in the path planning tasks of mobile robots, an Improved Crested Porcupine Optimizer (ICPO) based on chaotic mapping is proposed. The ICPO algorithm employs a three-step optimization process. First, it utilizes SPM, a piecewise linear chaotic initialization, to optimize the population thereby enhancing its diversity and global coverage. Second, the Cauchy Distribution Inverse Cumulative Operator is incorporated to prevent convergence to local optima and to accelerate the overall convergence rate. Finally, the Gaussian mutation is applied to strengthen ICPO’s local exploitation capabilities. Comparative analysis of five algorithms (PSO, DBO, GOOSE, CPO, and ICPO) is conducted using eight standard benchmark functions. Results demonstrate that ICPO achieves a faster convergence rate and superior convergence accuracy. Furthermore, in path planning experiments within 20 × 20 and 40 × 40 grid maps, ICPO reduced the path length by 4.53% and 8.99%, respectively, compared to the CPO algorithm.