Wireless Sensor-Actuator Network (WSAN) is a network embedded with a few powerful actuators on the basis of the original WSNs. Based on perception of the WSAN, those actuators can aggregate and process the obtained information in real time, and thus reduce the transmission of redundant information in the network. This helps to save energy consumption. However, due to time delay in data analysis and fusion in a WSAN, the service sensitive to time is provided with poor performance. Therefore, how to balance energy saving and time delay is the main problem in a WSAN. A time-sensitive WSAN network model is established in this paper. To realize the hierarchical structure of the network, the clustering algorithm is used to cluster the sensor nodes. The design of the network routing algorithm based on clustering is formed to the classical Traveling Salesman Problem (TSP). In our WSAN model, we propose the Shuffled Frog Leaping and Ant Colony Algorithm (SFL-ACA) node clustering algorithm. In some practical application scenarios, we proposed an improved scheme to further reduce the delay of network transmission. Based on the WSAN network simulation, we study the influence of different number of actuator nodes on the performance of WSANs, which therefore produces an effective number of actuator nodes deployment scheme in this paper.