Design optimization for a compliant mechanism based on computational intelligence method

摘要: Modeling and optimization for compliant mechanisms are challenging tasks thanks to an unclear kinematic merging among rigid flexible links. Hence, this paper develops a computational intelligence-based method modeling optimization. The proposed concerns about statistics, numerical simulation, intelligence, metaheuristics. A two degrees of freedom mechanism is investigated illustrate the effectiveness suggested intelligence method. First, datasets collected by simulations. Then, sensitivity design parameters analyzed analysis variance Taguchi technique. results employed separate few populations lightning attachment procedure (LAPO). Next, values output performances changed into in range from zero one through desirability function calculated become inputs fuzzy logic model, system single objective (SOF). Subsequently, SOF modeled using adaptive neuro-fuzzy inference (ANFIS). LAPO algorithm then utilized maximize SOF. revealed that example 3 best mechanism. In comparison with artificial techniques regression, show performance indexes ANFIS model (R2 close 1, MSE 10–4, RMSE 10–2) superior those multilayer perceptron, deep neural network, multiple-linear regression. Additionally, more effective than Taguchi-fuzzy logic, ANFIS-integrated teaching learning-based optimization, Jaya searching optimal determined displacement parasitic error 2.2109 mm 0.0028 mm, respectively.