In the context of low-carbon development, to reduce line losses in photovoltaic power plants and improve wiring efficiency, this paper proposes a cable optimization method that integrates the improved KICCA clustering algorithm with the SA-TS resource matching algorithm. For the photovoltaic array wiring problem, the improved KICCA algorithm enhances clustering accuracy and speed by employing ordered initialization of clustering centers (replacing random initialization), an extended Manhattan distance dissimilarity matrix (compatible with dual-cable endpoint selection), and a breadth-first neighbor search strategy. For the computational resource matching problem, the improved SA-TS algorithm is proposed by combining the global exploration of simulated annealing (SA) with the anti-repetition characteristics of tabu search (TS). Through resource classification quantification, pheromone weighting, and decision-making, as well as centralized/decentralized dual-mode load calculation, efficient resource scheduling is achieved. Experiments show that the algorithm converges after 23 to 38 iterations, achieving over 40% faster performance than traditional methods. The optimized solution reduces the voltage difference at the end nodes to 0V, significantly improving voltage consistency. In application tests on missile cable networks, the SA-TS algorithm achieved an automatic wiring length of only 3,700 mm, a 3.4% reduction compared to the manual solution, and reduced the design cycle from 20 days to 10 days, improving efficiency by 50%. In summary, this method optimizes cable paths through two-level clustering and combines intelligent resource matching to provide technical support for low-carbon construction of photovoltaic power plants, while verifying its universality in complex threedimensional spaces (such as missile cable laying).
