Optimal Arrangement of Power Cables in Ducts Using the Agamogenetic Algorithm

Duct bank installation, as a popular installation type for underground power cables, has been widely used. However, laying excessive number of ducts in a duct bank has given rise to the cable arrangement problem. Meanwhile, the current academic research on the cable arrangement problem is not mature. In this regard, this paper proposes a novel stochastic optimization algorithm namely the agamogenetic algorithm to solve the cable arrangement problem. The roulette strategy is applied in the mutation process of the proposed algorithm, leading to the development of two novel operators, namely the excitation operator and the interchange operator. From the general agamogenetic algorithm, this paper further designs its customized variants for the two-ends and single-end grounding modes. Subsequently, the concept of cross-bonded grounding mode is promoted, and a cross-bonded configuration beyond the same transmission circuit is proposed, namely the generalized cross-bonded grounding mode. Due to its requirement to simultaneously optimize the cable arrangement and the sheath cross-bonded structure, a specific oriented agamogenetic algorithm is also designed. Finally, a real-life duct bank installation is used as the case study, validating that the agamogenetic algorithm has higher convergence speed, accuracy, and stability compared to existing algorithms for optimizing cable arrangements in ducts. Simulation results also show that the loss or voltage of the optimal arrangement decreases by 12.24%, 50.66%, and 11.67% under the two-ends, single-end, and cross-bonded grounding modes, respectively, compared to the standard engineering arrangement.