Abstract The soft actuator has unlimited freedom, and can realize the grasping of irregular or fragile objects. A pneumatic soft actuator is designed in this paper, which consists of a series of symmetrically distributed pneumatic chambers. The soft actuator consists of a top expansion layer composed of 20 chambers and a bottom restraint layer. The Yeoh model and the finite element method (FEM) were employed to analyze the bending of the soft actuator. To improve the tracking performance, a PID strategy of the endocrine system (ES-PID) based on the neuroendocrine thyroid hormone regulation mechanism was constructed to tune the bending deformation. To achieve steady and precise force control, a bending/force switching control algorithm based on multi-sensor fusion is designed to realize high control accuracy. The bending/force switching control strategy mainly consists of two feedback loops for bending control, a force control, and a switching control. In the experimental process, the bending control is the primary control, and the force control is the primary control when the switching condition is satisfied. In order to prevent the occurrence of a false switching action, a switching factor is constructed using the method of successive comparison of neighboring values. Finally, several experiments are carried out to verify the effectiveness and feasibility of the control algorithm.