Thomas Hemker

Efficient Walking Speed Optimization of a Humanoid Robot

The development of optimized motions of humanoid robots that guarantee fast and also stable walking is an important task, especially in the context of autonomous soccer-playing robots in RoboCup. We present a walking motion optimization approach for the humanoid robot prototype HR18 which is equipped with a low-dimensional parameterized walking trajectory generator, joint motor controller and an internal stabilization. The robot is included as hardware-in-the-loop to define a low-dimensional black-box optimization problem. In contrast to previously performed walking optimization approaches, we apply a sequential surrogate optimization approach using stochastic approximation of the underlying objective function and sequential quadratic programming to search for a fast and stable walking motion. This is done under the conditions that only a small number of physical walking experiments should have to be carried out during the online optimization process. For the identified walking motion for the considered 55 cm tall humanoid robot, we measured a forward walking speed of more than 30 cm s -1 . With a modified version of the robot, even more than 40 cm s -1 could be achieved in permanent operation.

Applicability of surrogates to improve efficiency of particle swarm optimization for simulation-based problems

Particle swarm optimization (PSO) is a population-based, heuristic technique based on social behaviour that performs well on a variety of problems including those with non-convex, non-smooth objective functions with multiple minima. However, the method can be computationally expensive in that a large number of function calls is required. This is a drawback when evaluations depend on an off-the-shelf simulation program, which is often the case in engineering applications. An algorithm is proposed which incorporates surrogates as a stand-in for the expensive objective function, within the PSO framework. Numerical results are presented on standard benchmarking problems and a simulation-based hydrology application to show that this hybrid can improve efficiency. A comparison is made between the application of a global PSO and a standard PSO to the same formulations with surrogates. Finally, data profiles, probability of success, and a measure of the signal-to-noise ratio of the the objective function are used to assess the use of a surrogate.

DERIVATIVE-FREE OPTIMIZATION METHODS FOR HANDLING FIXED COSTS IN OPTIMAL GROUNDWATER REMEDIATION DESIGN

We consider a hydraulic capture application for water resources management that includes a fixed installation cost in addition to operating costs. The result is a simulationbased, nonlinear, mixed-integer optimization problem. The motivation is that our preliminary studies have shown that convergence to an unsatisfactory, local minimum with many wells operating at low pumping rates is common when the fixed cost is ignored. Such optimization tasks are not unique to subsurface management, rather efficient simulation-based methods are needed in the whole field of computational engineering. All the approaches used below do not need the gradient of the objective function, only function values for minimization. In one approach, we bypass including the number of wells as a decision variable by defining an inactive-well threshold. In another approach, we use penalty coefficients proposed in the literature to transform the discontinuous problem into a continuous one. For the two above formulations, we use the implicit filtering algorithm. In the third approach, we introduce a mixed-integer problem formulation and use an iterative stochastic modeling technique to build surrogate functions that approximate the objective function. With this new procedure the use of a branch-and-bound technique becomes possible to solve the mixed-integer problem in contrast to methods working directly on the simulation results, which impedes relaxation of integer variables. We present promising numerical results on the benchmarking problem and point the way towards improvement and future work.

Analysis of Industrial Robot Structure and Milling Process Interaction for Path Manipulation

Industrial robots are used in a great variety of applications for handling, welding, assembling and milling operations. Especially for machining operations, industrial robots represent a cost-saving and flexible alternative compared to standard machine tools. Reduced pose and path accuracy, especially under process force load due to the high mechanical compliance, restrict the use of industrial robots for machining applications with high accuracy requirements. In this chapter, a method is presented to predict and compensate path deviation of robots resulting from process forces. A process force simulation based on a material removal calculation is presented. Furthermore, a rigid multi-body dynamic system’s model of the robot is extended by joint elasticities and tilting effects, which are modeled by spring-damper-models at actuated and additional virtual axes. By coupling the removal simulation with the robot model the interaction of the milling process with the robot structure can be analyzed by evaluating the path deviation and surface structure. With the knowledge of interaction along the milling path a general model-based path correction strategy is introduced to significantly improve accuracy in milling operations.

A novel self-calibration method for industrial robots incorporating geometric and nongeometric effects

We propose a novel problem formulation for the solution of the nonlinear least squares appearing in the static calibration of the parameters of an industrial robot with rotational joints. A further contribution of the presented work is the use of an extended forward kinematic model incorporating both geometric and nongeometric parameters. These novelties facilitate the use of industrial robots as measurement systems. Both the automobile and the general industry are interested in the use of highly accurate robots that can be deployed as measurement instruments, e.g. for noticing incorrectly welded points early in the production process. Instead of absolute measurement data, we make use of relative measurement data by means of a camera attached to the robot flange. In tests based on data generated from simulation of a real experimental setup, we obtain after calibration absolute accuracies better than plusmn 100 mum. This work lays the foundations for a cost-minimal and effective realization of a robot as a measurement instrument.

Towards the deployment of industrial robots as measurement instruments - An extended forward kinematic model incorporating geometric and nongeometric effects

In the area of mounting and spot-welding of body-in-white, absolutely accurate robots are installed as measuring instruments, replacing expensive coordinate and other external measuring machines. Measurement technologies based on industrial robots play an increasingly important role. Such applications require highly accurate robots. Prior to deployment of highly accurate robot, however, it needs to be ensured that the implemented robot model fits the real model. Robot calibration can offer a significant opportunity to improve the positioning accuracy and to cut production costs. Existing calibration approaches fail to capture geometric and elastic effects occurring in the robot forward kinematics. Therefore, in this work an extended forward kinematic model incorporating both geometric and elastic effects has been developed in which the positioning accuracy of a manipulator, with or without an accurate internal robot model in the robot controller, is improved.

Mixed-Integer Nonlinear Design Optimization of a Superconductive Magnet With Surrogate Functions

The numerical optimization of continuous parameters in electrotechnical design using electromagnetic field simulation is already standard. When integer-valued variables are involved, the complexity of the optimization problem rises drastically. In this paper, we describe a new sequential surrogate optimization approach for simulation-based mixed-integer nonlinear programming (NLP) problems. We apply the method for the optimization of combined integer- and real-valued geometrical parameters of the coils of a superconductive magnet.

Self-calibration for industrial robots with rotational joints

We suggest a novel self-calibration method for industrial robots. Measurement technologies based on industrial robots play an increasingly important role. Existing approaches are disadvantageous because of the use of additional external measuring systems and the time-consuming and cost-intensive determination of the robot base. We propose a procedure for a simple experimental setup including one CCD camera. A further novelty is the use of an extended forward kinematic model incorporating both geometric and nongeometric parameters. Both novelties facilitate the use of industrial robots as measuring systems. This work lays the foundations for a cost-minimal and effective deployment of robots as measuring instruments.