The main research aim of this thesis is to design the system architecture of the Stewart Platform in order to make the related operational processes being more efficient and convenient. As the result, we developed a graphical user interface for the goal of effective adjustment of corresponding control parameters, which is implemented by I/O signal cards to transmit the data files between the control console and the platform. In other words, within the limited working space, we can control the motion cueing of the platform arbitrarily and immediately. The Functional Neural Fuzzy Controller (FNFC) was developed by our Lab, while the corresponding algorithms consist of the structure and parameter learning processes. The process of structure learning is based on the entropy measurement to decide whether there should be added one more fuzzy rule, and the backpropagation method is used in the parameter learning to adjust all free parameters of the neural network. Comparing to the existing PID controllers for evaluation of motion platforms, our approach could provide better performance, reduce the corresponding complexity, and minimize the error between actual length and commanded length of the platform's actuators. In order to maximize the working space within the realistic limited range and to simulate the large scale motions of the relative acceleration and angular velocity and to make experienced pilots feeling more realistically under their operations, we applied the algorithm of the washout filter to immediately and smoothly perform each commanded motion cues on our platform. Lastly, we compared our proposed approach with other methods to prove the effectiveness of our neural architecture and the corresponding algorithms.