| On account of the limitation of dimensions,traditional workspace of parallel mechanism is always separated as position workspace and orientation workspace.In order to deal with this low visualization workspace representation problem,a novel multi-dimensional workspace representation method for parallel mechanism is proposed,which can realize the combination among different dimensional variables.The multi-dimensional representation method can be used to represent the workspace and performance space of parallel mechanism.Based on the method,a group of novel workspace performance indexes for parallel mechanism is proposed.A novel task-oriented structural design method for parallel mechanism is also presented.Multi-dimensional workspace of parallel mechanism can be constructed by adding figure dimension,which is achieved through extending point cloud among traditional 2D and 3D workspace diagrams.While multi-dimensional performance space of parallel mechanism can be constructed by applying performance indexes as additional dimensional variable.By mixing multiple dimension and measurement unit variables,the motion information which includes translational,rotational movements and stiffness et al can be combined into one multi-dimensional workspace or performance space diagram.The visualization level of diagrams can be improved.Based on the features of multi-dimensional workspace and performance space,a novel mathematical model of mechanism performance can be proposed.Two typical parallel mechanisms are presented and kinematics is analyzed.By comparing workspace and performance space with traditional method,the correctness and advantage of multi-dimensional representation method have been verified.Optimization also has been carried out by applying novel performance indexes as objects,which verified the applicability of multi-dimensional performance indexes.Based on the multi-dimensional representation method,a visible task-oriented structural design method for parallel mechanism is proposed.First,suitable motion branches which satisfy task DOF should be carried out.Then,according to the shape of task workspace and overall size of mechanism,standardization is performed on task workspace.Standardized task workspace is applied in the following design step.Size range of the mechanism is determined by workspace matching and neural network.Optimization also can be carried out by genetic algorithm,which allows the mechanism owns maximum performance while fitting task requirements.The task-oriented structural design method is applied to constructing planer and spatial parallel mechanisms based on assumed tasks.Task fitting verification is presented by comparing multi-dimensional diagrams among different mechanisms.Moreover,a kind of parallel mechanism with variable workspace has been carried out using the structural design method.Size ranges of the mechanism can be preliminarily determined by standardized task performance space,which satisfies the volume of task performance space.More accurate size range is determined by BP neural network,which fits all task requirements.Optimization has been carried out by genetic algorithm as well.Finally,the design and usage of multi-dimensional representation programmed algorithm is given,an example case also has been presented. |
PDF Full Text Request