The mechanism type plays a decisive role in the mechanical performance of robotic manipulators. Feasible mechanism types can be obtained by applying appropriate type synthesis theory, but there is still a lack of effective and efficient methods for the optimum selection among different types of mechanism candidates. This paper presents a new strategy for the purpose of optimum mechanism type selection based on the modified particle swarm optimization method. The concept of sub-swarm is introduced to represent the different mechanisms generated by the type synthesis, and a competitive mechanism is employed between the sub-swarms to reassign their population size according to the relative performances of the mechanism candidates to implement the optimization. Combining with a modular modeling approach for fast calculation of the performance index of the potential candidates, the proposed method is applied to determine the optimum mechanism type among the potential candidates for the desired manipulator. The effectiveness and efficiency of the proposed method is demonstrated through a case study on the optimum selection of mechanism type of a heavy manipulator where six feasible candidates are considered with force capability as the specific performance index. The optimization result shows that the fitness of the optimum mechanism type for the considered heavy manipulator can be up to 0.578 5. This research provides the instruction in optimum selection of mechanism types for robotic manipulators.
This paper proposes an approach to evaluate the performance of robot manipulator from the view of energy analysis. Based on the dynamics analysis of the manipulator, the Energy Distribution Index （EDI） is defined to depict the energy increment contribution of its subsystem to the whole manipulator. EDI is applied to the evaluation of the buffeting capability of the manipulator working under unpredictable and heavy external loads. A series of buffering indices, the Static Buffering Index （SBI）, Kineto-Static Buffering Index （KBI）, Dynamic Buffering Index （DBI）, and Global Buffering Index （GBI） are proposed to evaluate the buffering capability under different conditions. In order to acquire higher calculation accuracy, the general stiffness mapping of manipulators considering the actuator stiffness, inertia of the manipulator, damping, as well as elasticity of linkages is developed. Three different robot manipulators are studied as evaluation cases, in which the buffering structures are mechanism with variable topology, linear springs, and the elasticity of linkages respectively. The case studies show that the indices based on energy analysis have the advantage of coordinate free and are effective for buffering capability evaluation.
