Fairing Algorithm of T-spline Surface Based on the Improved Sigmoid Function

T-spline surfaces play important role in the integration of CAD and CAE, the quality of the surface infects not only the design of applications but also the analysis in simulations. So fairing is a necessary connection after designing and before analysis inherently. The Sigmoid function is usually adopted in a neural network as an activation function and possess continuous and smooth property, we introduce it into surface fairness taking for the weight of the fairing function, and the parameter of the improved Sigmoid function is determined by considering space angle and fitting previous approach. Several experiments are exhibited to indicate the validity of the proposed method, and it reaches a competitive result while less procedure and more stability compared with the existing algorithm.