SSMLoRA: Enhancing Low-Rank Adaptation with State Space Model