Abstract: Abstract: In order to improve production efficiency, reduce production cost, cope with the growing labor shortage, and improve the uniformity of feeding, the use of smart equipment is an inevitable trend. Due to the limitations of the industry property, the work of aquaculture industry is simple, tedious, and heavy. Using smart equipment substituted for human, it is the most suitable and competitive. In order to carry equipment on water surface and to move, it needs mobile operating platform. Paddlewheel vehicle is in line with the requirements. This paper studied the sailing control of paddlewheel vehicle. The paddlewheel vehicle could go on forward, which relied on both sides of the wheel propulsion device - paddlewheel. Due to the inherent characteristics of the paddlewheel vehicle structure, there was a strong coupling between the speed and the heading. It could cause speed fluctuations in the movement of the steering process. To achieve the designed speed and heading, both sides of the paddlewheel rotating speed will fluctuate dramatically. This situation reduced sailing stability of paddlewheel vehicle, wasted limited energy, shortened total route mileage, and influenced the operation of carrying device on the ship. To overcome this problem, the control system was analyzed and pointed out the defects, and then the improved paddlewheel vehicle control system was given, which had the advantage of decoupling effect between the heading and the speed of paddlewheel vehicle. The improved control system was important to make paddlewheel vehicle have better sailing stability. After that, according to the Fossen theory, the heading system dynamic response model was acquired from the hydrodynamic equations of paddlewheel vehicle. Next, on the base of analyzing work situations of paddlewheel vehicle, the paper proposed the method of multi-mode adaptive control. It treated the paddlewheel vehicle's sailing straight and steering movements with different methods. In order to improve the dynamic response effect and anti-jamming capability of paddle vehicle, the fuzzy logics control heading system was used for paddlewheel vehicle. In order to verify the validity of the controller and control algorithms, the dynamic response model was constructed using the MATLAB software. Then, the simulation was conducted by using computer with the designed algorithms of the multi-mode adaptive control. At the same time, the simulation of PD (proportion differentiation) control was carried out for comparing. Then, the experiment was carried out on water surface in outdoors. GPS (global position system) and inertial navigation equipment were installed on paddlewheel vehicle, and they were used to measure and collect the data of real-time tracking and heading angle, which were transmitted to the host computer and recorded through serial communication device. By comparing the simulation and experimental results, it was found that using multi-mode adaptive control could effectively control the paddlewheel vehicle's stability on the basis of the improved control system solutions. Results showed that the speed overshoot was no more than 5%, the steady-state error within 3%, and the heading error within 3° when straight sailing. Both methods of multi-mode adaptive control and PD control were carried out on the paddlewheel vehicle heading. Computer simulation and experiments were used to test and compare control effect. The results of simulation and experiment prove that the designed controller can achieve the decoupling, and the multi-mode adaptive control method can control the paddlewheel vehicle course, which is better than the PD control.