B.K. Rout

Optimal manipulator parameter tolerance selection using evolutionary optimization technique

Robot system designers often face the challenge of selecting optimal parameter tolerances of a manipulator, which delivers optimal performance. This paper presents an approach to simulate the performance of manipulator and evolutionary optimization method to select optimal parameter tolerance. To determine optimal parameter tolerance, genetic algorithm, and differential evolution, optimization techniques have been used. The objective function maximizes SN Ratio, while manipulator performs a task. As differential evolution and GA are best suited for solving deterministic optimization problems, to handle performance of manipulator, a hybrid technique is proposed. The evolutionary optimization techniques are coupled with orthogonal array used in the Taguchi method to get optimal solution. The hybrid technique is illustrated by an example and concluded that it is best suited for manipulator parameter tolerance design. It is also observed that differential evolution technique converges quickly and require significantly less number of functional evaluations.

Optimal design of manipulator parameter using evolutionary optimization techniques

A robot must have high positioning accuracy and repeatability for precise applications. However, variations in performance are observed due to the effect of uncertainty in design and process parameters. So far, there has been no attempt to optimize the design parameters of manipulator by which performance variations will be minimum. A modification in differential evolution optimization technique is proposed to incorporate the effect of noises in the optimization process and obtain the optimal design of manipulator, which is insensitive to noises. This approach has been illustrated by selecting optimal parameter of 2-DOF RR planar manipulator and 4-DOF SCARA manipulator. The performance of proposed approach has been compared with genetic algorithm with similar modifications. It is observed that the optimal results are obtained with lesser computations in case of differential evolution technique. This approach is a viable alternative for costly prototype testing, where only kinematic and dynamic models of manipulator are dealt with.

Investigation on parametric sensitivity of topologically optimized structures

Topology optimization is a powerful method of material minimization in structural design problems. The obtained topology and the compliance values by this method are very sensitive to each of the input parameters such as, applied force, volume fraction, dimensions, and support-rigidity. In real-life situations, these parameters may vary due to material uncertainty, manufacturing imperfections, and operating conditions. Hence, the topology obtained during the conceptual design phase may not suffice the actual working condition. Thus, it is desirable to explore individual and the combined effects of the parametric variations and uncertainties. This study describes a systematic approach utilized to investigate the effect of different input parameters on compliance values along with material and load uncertainties for a topologically optimized structure. In this paper, applied force, volume fraction, and aspect ratio of the domain are treated as input parameters and their effects are analyzed. Proposed work modifies the solid isotropic microstructure with penalization method to incorporate the effect of uncertainties and uses design of experiments approach to investigate statistically significant input parameters. Four different benchmark problems available in the literature are analyzed and the results are obtained for aforesaid input parameters along with uncertainties. Results obtained from this investigation will help designers/practitioners to select suitable input parameters combination to achieve targeted compliance.

Tolerance range selection of topologically optimized structures with the effects of uncertainties of manufacturing process

Robust design methods for topology optimization have received significant attention from researchers in recent years. There are various attempts in past to handle the manufacturing uncertainty of the topologically optimized components. In the present work, same issue is dealt by introducing a method of probabilistic distribution of material. Here, help from Karhunen–Loeve expansion of stochastic process, coupled with Monte Carlo method is taken. The proposed method retains the uncertainty characteristic of the specific manufacturing processes such as etching, e-beam lithography, laser micro machining, and milling. Hence, this method offers larger flexibility to the designer. The simulation for manufacturing uncertainty is utilized to determine the optimal tolerance range of the factors for robust and targeted performance, which is almost nonexistent in the literature. The methodology for tolerance range selection is capable to incorporate uncertainty of multiple factors simultaneously. In this method cross array design of experiment approach is used to analyze the effect of tolerance of each factor. The overall process for manufacturing uncertainty and tolerance range selection is illustrated using four benchmark problems. The chosen factors for considered structural problems are force, elasticity, volume fraction, and aspect ratio. Various combinations of tolerance ranges are used to simulate the performance of the optimized structure, which is expressed in terms of robustness and targeted values of compliance, and maximum deflection. Based on the simulated results of signal-to-noise ratio and mean values of performance, the combinations of tolerance range are suggested that gives a high level of robustness or targeted performance accuracy. To indicate the uniqueness of proposed approach, the obtained response for performances is compared with already available response for performance in literature for generalized approach. Current work is advantageous compared to usual robust design, and provides the performance for a specific scenario at each possible combination of tolerance ranges.

Optimal manipulator parameter selection using evolutionary optimization technique

This article discusses an approach to selecting the optimal design parameters of a manipulator for minimum performance variations. The variations in desired performance of the manipulator are attributed to the effect of uncertainty in design, process and noise factors. To incorporate the effect of uncertainties in design parameters, a modification in evolutionary optimization approach is proposed. A worst-case approach has been used to model the uncertainties and simulate the performance of the manipulator while performing a task. The design parameters obtained from this optimization process are insensitive to the effect of uncertainties. The proposed approach has been illustrated with the help of a 2-DOF RR planar manipulator. This approach has also been implemented in a genetic algorithm to compare the computational advantage of the proposed modification in the differential evolution technique. The optimal design parameters are observed to be different for different tasks and trajectories in the workspace.

Optimal manipulator tolerance design using hybrid evolutionary optimization technique

There is a need to select optimal parameter tolerance of manipulator to reach an economic balance between the desired performance and its manufacturing cost. However, selection of optimal parameter tolerances of manipulator is a challenging task. Present paper discusses an offline approach to incorporate effect of noise in simulation of performance and handle its effect in optimization process of parameters tolerances. To determine optimal parameter tolerances, a hybrid evolutionary optimization technique has been used. The hybrid is formed between differential evolution optimization technique and orthogonal array used in design of experiments technique. Proposed technique has been illustrated by selecting optimal tolerances of 2-DOF RR planar manipulator. It has been observed that the methodology is a viable alternative to the costly prototype testing, where only mathematical models are dealt with.