Abstract: As the core technology of the robot controller, the motion control algorithm has an important influence on the robot motion performance including stability, reliability, and rapidity. However, the application of motion control algorithms usually faces the problems of poor generality, complex embedded algorithm, and long design cycles. Aiming at these problems, this study combines the complex flexible S-shaped acceleration/deceleration (ACC/DEC) motion control algorithm and proposes a model-based hardware-software collaboration design method, which can greatly shorten the design cycle of the robot motion control system and improve the development efficiency. By modeling the flexible motion control algorithm and establishing a set of interface parameter list easily calculated, the proposed algorithm can adaptively change the motion speed planning according to the parameters in the list such that the flexibility of application can be improved. The model design and simulation test for flexible motion control algorithms are completed in Simulink platform. The Math Works toolbox is utilized to automatically generate the embedded C code and programmable logic IP cores for the software and hardware models. Finally, the proposed algorithm function is realized in the motion controller based on zynq-7000. It is verified via the simulation results that the designed model can achieve the effect of the S-shaped ACC/DEC algorithm and the speed profile has great flexible characteristics. Moreover, the ACC/DEC algorithm deployed on the Zynq-7000 is in agreement with the simulation results. Hence, the model-based software and hardware co-design method has important application value in the field of personalized robots.