To address the challenges of excessive pilot overhead and limited adaptability in cascade channel estimation for reconfigurable intelligent surface （RIS） assisted wireless communication systems，this paper proposes an improved whale optimization algorithm integrated with a dual-structure sparse stagewise weak orthogonal matching pursuit algorithm （IWOA-DS-SWOMP）. The framework employs an adaptive threshold-controlled SWOMP mechanism to iteratively select multiple highly correlated atoms for constructing atomic support sets， while the atomic selection threshold via IWOA is dynamically optimized to adapt to real-time channel variations. This dual optimization strategy enhances atomic support set extraction accuracy， improves channel estimation precision， and reduces algorithm runtime. Simulation results demonstrate that the proposed scheme achieves superior normalized mean square error （NMSE） performance compared to conventional RIS cascade channel estimation methods， attaining higher channel estimation accuracy with reduced pilot overhead while exhibiting enhanced adaptability and robustness under diverse channel conditions.