Pareto optimization-based trade-off between the visual resolvability and the intersection space of view of binocular microscopic vision system for microassembly systems

Visual resolvability and intersection space of view of a binocular microscopic vision system (MVS) in microassembly systems have influences on each other via analyzing theoretically. According to requirements of various microassembly tasks, it is an explicit problem how to find the best compromise between the visual resolvability of binocular MVS and its intersection space of view in three-dimensional space. A Pareto optimization-based method to make the trade-off between the visual resolvability and the intersection space of view of binocular MVS with relevant fixed parameters in microassembly systems is proposed. Through modeling and analyzing on the visual resolvability and intersection space of view of a binocular MVS, the topology and optical magnifications of two MVSs are considered as the key parameters to optimize its physical properties. A multiobjection optimization model based on a Pareto optimization method is formulated and used to find the Pareto optimal cure, namely the Pareto front, which represents the best trade-off between the visual resolvability and the intersection space of view of a binocular MVS. On these bases, the physical properties of a binocular MVS with structural parameters corresponding to the optimization solutions are numerically simulated. Accordingly, a binocular MVS with different parameters is established on an automatic microassembly system with a single manipulator with five degree-of-freedom and experimentally tested. The simulation and experimental results indicate that all of the optimal solutions, which locate on the Pareto front, are the best trade-off between the visual resolvability and the intersection space of view of a binocular MVS. According to the Pareto front, physical parameters of a binocular MVS can be chosen to get the optimal visual resolvability with a certain intersection space of view and vice versa.