A coverage hole is a problem that cannot be completely avoided in three-dimensional hybrid wireless sensor networks. It can lead to hindrances in monitoring tasks and adversely affect network performance. To address the problem of coverage holes caused by the uneven initial deployment of the network and node damage during operation, we propose a distributed hole detection and multi-objective optimization emperor penguin repair algorithm (DHD-MEPO). In the detection phase, the monitoring region is zoned as units according to the quantity of nodes and the sensing range, and static nodes use the sum-of-weights method to campaign for group nodes on their terms, determining the location of holes by calculating the coverage of each cell. In the repair phase, the set of repair nodes is determined by calculating the mobile node coverage redundancy. Based on the characteristics of complex environments, the regions of high hole levels are prioritized. Moreover, the residual energy homogeneity of nodes is considered for the design of multi-objective functions. A lens-imaging mapping learning strategy is introduced to perturb the location of repair nodes for the optimization of the emperor penguin algorithm. Experimental results illustrate that the DHD-MEPO, compared with the C-CICHH, 3D-VPCA, RA, EMSCOLER, and IERP algorithms, can balance the uniformity of the residual energy of each node while satisfying the network coverage requirements and network connectivity, which effectively improves the network coverage performance.