Among the various industrial robots, the palletizing robot dis receiving particular attention due to its higher productivity in the industry along with technological progress. The palletizing robot is mainly used for logistics automation such as product transfer between processes in automated line. The palletizing robot is usually performs pick-and-place operation by applying mechanical restrictions using three parallelogram link mechanisms, also the servo motor can be placed lower than the overall center of gravity of the robot frame so that it has smoother operation and higher dynamic characteristics.
In initial designing of such robots, the servo motors should be properly selected in advance. However, dynamic analysis to calculate the driving torque normally requires a lot of time and effort. To facilliate such analysis processes, this study evaluates the driving torque of the loading and unloading using ADAMS, a dynamic analysis software, then analyzes the specifications of the initially selected servo motors and reducers. In this process, acceleration torque and gravity torque are also obtained.
The dynamic analysis is performed on the basis of the assumption that the links and frames of the palletizing robot are rigid. However, the links and frames can be actually considered as flexible body. In consideration with such flexibility, this study carries out flexible multibody dynamic analysis using ANSYS and ADAMS to propose a practical method for selecting the motors and reducers. In addition, the results of flexible multibody dynamic analysis are compared with the ones obtained from the rigid body case.
In designing of a palletizing robot, the overall weight of the link and frame should be minimized on the condition that structural stability is satisfied. For the purpose, this study carries out weight minimization with a genetic algorithm. As the previous step to the optimization, the link with the highest maximum stress and the frame that can be reduced in weight are selected through structural analysis. The dimensional parameters of the selected link and frame are also defined for use as the indivisuals of the genetic algorithm which is implemented using MATLAB. The implemented genetic algorithm evolves into the optimal indivisuals using the genetic operators selected through testing, such as a linear rank reproduction operator, an arithmetic crossover operator, and a dynamic mutation operator. By using elite strategy, the extinction of an optimal individual is prevented during the evolution process. Also, the fitness function defined in the maximization form of the evaluation function is defined to implement the genetic algorithm. Here, the evaluation function is defined with both the objective function and penalty function. The genetic algorithm requires structural analysis to obtain an evaluation function that reflects the penalty function as a limiting condition of maximum stress along with an objective function that minimizes the weight according to changes in dimensional parameters. In this study, ANSYS Workbench is used for structural analysis, and the genetic algorithm was implemented in MATLAB. The optimization simulation is performed by co-simulation of ANSYS Workbench and MATLAB. As a result of the optimization, the total weight of the link and frame is reduced by about 57% compared to the initial model.
In this study, using the proposed flexible body dynamic analysis instead of the rigid body dynamic analysis is confirmed to be more effective in selecting the motors and reducers when the acceleration and deceleration time is short and the effect of gravity is large. Also, the weight optimization of link and frame is performed using a genetic algorithm through co-simulation of ANSYS Workbench and MATLAB.