Evolutionary learning acquisition of trajectory of a rotary crane using genetic algorithm

Abstract

This paper proposes an evolutionary learning acquisition method of the trajectory of a rotary crane for the residual vibration reduction by use of the dynamic analysis of the rotary crane and the genetic algorithm. The rotary crane is driven by only the rotational motion of a boom. In the design of the evolutionary learning acquisition system, the equation of motion of the rotary crane is obtained, and the trajectory of the rotational angle of the boom is depicted with sixth-order polynomials. Then, considering the boundary and constraint conditions, the trajectory are expressed by two genes. Solving the equation of motion of the load corresponding to the trajectory of the boom, the residual vibrations of the load are computed as a function of the chromosome with the two genes, and a fitness function of the genetic algorithm for the residual vibration reduction of the load is defined. Furthermore, the evolutionary learning acquisition algorithm of the trajectory of the boom for the residual vibration reduction of the load is constructed. Moreover, it is confirmed numerically and experimentally that the residual vibrations of the load can be extremely reduced by the present evolutionary learning acquisition method.