Complete Coverage Path Planning (CCPP) in unstructured environments presents a critical challenge due to the necessity of balancing high coverage efficiency and path accuracy. Traditional Particle Swarm Optimization (PSO) is effective but tends to converge prematurely to local optima in complex scenarios. The paper proposes a Comprehensive Improved Particle Swarm Optimization (CIPSO) algorithm, which enhances conventional PSO through three targeted modifications to tackle the challenge in scenarios involving randomly distributed stains: (1) Dynamic merging of redundant path points reduces excessive revisitation; (2) A particle evaluation parameter mechanism is presented to adjusts learning factors to alleviate premature convergence; and (3) Hybridization with genetic algorithm strategies enhances global exploration capabilities. Experimental evaluations were performed in an 80 cm × 100 cm blackboard with randomly distributed stains, comparing the proposed CIPSO against standard PSO and dynamic-weight PSO variants. Results indicate that CIPSO achieves a 23% reduction in comprehensive costs compared to standard PSO and outperforms dynamic-weight IPSO by 15%. Furthermore, CIPSO shows strong stability across randomized trials, effectively balancing efficiency and accuracy in unstructured path planning, offering a promising CCPP optimization framework.