Abstract—This paper presents the results of a newly developed hybrid Simulated Annealing (HSA) for the optimization of mechanism for function generation problem. A hybrid optimization method based on the fusion of the Simulated annealing (SA) and Marquardt Search (MS), gradient search based method nonlinear regression algorithm, in which the SA is embedded the MS to enhance its search capability. The paper describes a new hybrid optimization method that combines advantages of both global and local search. A brief overview of HSA is presented and applied to dimensional synthesis of a planer four bar mechanism. The optimization is carried out to minimize the objective function formulated from the structural error at the accuracy points. It is employed to determine the optimal values for the design variables that best satisfy the desired objectives of the problem. Simulation results demonstrate the remarkable advantages of our approach in achieving the diverse optimal solutions and improved converge speed. The applicability of HSA algorithm is illustrated by solving two nonlinear function generation problems. It produces accurate and acceptable solution in all cases.
Index Terms—Function generation, Simulated annealing, Marquardt search, Hybrid algorithm, and four bar mechanism.
[PDF]
Cite: M. K. Sonpimple, P. M. Bapat, J. P. Modak and S. R. Pimpalapure, "A Hybrid Simulated Annealing Algorithm for Mechanism Synthesis with N-Accuracy Points,"
International Journal of Engineering and Technology vol. 2, no. 4, pp. 367-373, 2010.
