Showing papers by "M Maarten Steinbuch published in 2013"

Optimal gear shift strategies for fuel economy and driveability

Abstract: This paper aims at designing optimal gear shift strategies for conventional passenger vehicles equipped with discrete ratio transmissions. In order to study quantitatively an optimal trade-off between the fuel economy and the driveability, the vehicle driveability is addressed in a fuel-optimal gear shift algorithm based on dynamic programming by three methods: method 1, weighted inverse of power reserve; method 2, constant power reserve; method 3, variable power reserve. Furthermore, another method based on stochastic dynamic programming is proposed to derive an optimal gear shift strategy over a number of driving cycles in an average sense, hence taking into account the vehicle driveability. In contrast with the dynamic-programming-based strategy, the obtained gear shift strategy based on stochastic dynamic programming is real time implementable. A comparative analysis of all proposed gear shift methods is given in terms of the improvements in the fuel economy and the driveability. The variable-power-re...

Learning intentions for improved human motion prediction

Abstract: For many tasks, robots need to operate in human populated environments. Human motion prediction therefore is gaining importance. The concept of social forces defines virtual repelling and attracting forces from and to obstacles and points of interest. These social forces can then be used to model typical human movements given an environment and a person's intention. This work shows how such models can exploit typical motion patterns summarized by growing hidden Markov models (GHMMs) that can be learned from data online and without human intervention. An extensive series of experiments shows that exploiting the intended position estimated using a GHMM within a social forces based motion model yields a significant performance gain in comparison with the standard constant velocity-based models.

Experimental validation of a dynamic waste heat recovery system model for control purposes

Abstract: This paper presents the identification and validation of a dynamic Waste Heat Recovery (WHR) system model. Driven by upcoming CO2 emission targets and increasing fuel costs, engine exhaust gas heat utilization has recently attracted much attention to improve fuel efficiency, especially for heavy-duty automotive applications. In this study, we focus on a Euro-VI heavy-duty diesel engine, which is equipped with a Waste Heat Recovery system based on an Organic Rankine Cycle. The applied model, which combines first principle modelling with stationary component models, covers the two-phase flow behavior and the effect of control inputs. Furthermore, it describes the interaction with the engine on both gas and drivetrain side. Using engine dynamometer measurements, an optimal fit of unknown model parameters is determined for stationary operating points. From model validation, it is concluded that the identified model shows good accuracy in steady-state and can reasonably capture the most important dynamics over a wide range of operating conditions. The resulting real-time model is suitable for model-based control. Copyright © 2013 SAE International.

Deformable mirrors with thermo-mechanical actuators for extreme ultraviolet lithography: Design, realization and validation

Abstract: In lithographic illumination systems, a nonuniform light distribution causes local deformations on the mirrors used. Active mirrors are a solution to correct these deformations by reshaping the surface. This paper presents the deformation of a mirror with thermo-mechanical actuators placed perpendicular to the surface. Two deformable mirrors are modeled, realized and validated: one with seven and one with 19 actuators. By placing the actuators on a thin back plate, the force loop is localized and therefore a lower actuator coupling is achieved. The thermo-mechanical actuators are free from mechanical hysteresis and therefore have a high position resolution with high reproducibility. Extensive Finite Element Analysis is done, to maximize actuator stroke and minimize input power. The mirrors are tested and validated with interferometer surface measurements and thermocouple temperature measurements. A mirror deflection of 0.68 nm/K is realized and no hysteresis is observed. Thermal step responses are fitted and both heating and cooling characteristic time constants are 2.5 s. The thermal actuator coupling from an energized actuator to its direct neighbor is 6.0%. The total actuator coupling is approximated around 10%, based on the good agreement between simulated and measured inter-actuator stroke.

Heat exchanger modeling and identification for control of Waste Heat Recovery systems in diesel engines

Abstract: To meet future CO2 emission targets, Waste Heat Recovery systems have recently attracted much attention for automotive applications, especially for long haul trucks. This paper focuses on the development of a dynamic counter-flow heat exchanger model for control purposes. The model captures the dynamic phenomena of two-phase fluid flow using the mass and energy balance equations. While most of the studies use chemical libraries to retrieve the working fluid properties, in this model mathematical equations are derived. Compared to other evaporator models, the proposed model is validated on data from a complete engine platform. Experiments are done on a state-of-the-art Euro-VI heavy-duty diesel engine, which is equipped with a Waste Heat Recovery system. For transient conditions over a wide range of operating points, simulation results show good agreement in comparison with experimental data. This makes the model suitable for real-time simulations, diagnostics and control algorithm designs.

Active Object Search Exploiting Probabilistic Object-Object Relations

Abstract: This paper proposes a probabilistic object-object relation based approach for an active object search. An important role of mobile robots will be to perform object-related tasks and active object search strategies deal with the non-trivial task of finding an object in unstructured and dynamically changing environments. This work builds further upon an existing approach exploiting probabilistic object-room relations for selecting the room in which an object is expected to be. Learnt object-object relations allow to search for objects inside a room via a chain of intermediate objects. Simulations have been performed to investigate the effect of the camera quality on path length and failure rate. Furthermore, a comparison is made with a benchmark algorithm based the same prior knowledge but without using a chain of intermediate objects. An experiment shows the potential of the proposed approach on the AMIGO robot.

Design and control of high tech systems

Abstract: Advanced motion systems like pick-and-place machines used in the semiconductor industry challenge the frontiers of mechatronic design, systems and control theory and practice. In the design phase, control oriented design of the electro-mechanics is necessary in order to achieve the tight performance specifications. Once realized, and since experimentation is fast, a machine in the loop procedure can be explored to close the design loop from experiment, experimental model building, model based control design, implementation and performance evaluation. Nevertheless, reliable numerical tools are required to meet the challenges posed with respect to dimensionality and model complexity. Extension of linear modeling techniques towards some classes of nonlinear systems is relevant for improved control of specific motion systems, such as with friction. Finally, medical robotics can greatly benefit from the experiences in high tech motion systems, and an eye surgical robot with haptics will be shown as an example. Other challenging applications in need for advanced design, modeling and control are fuel-efficient vehicles, including ultra-clean engines, vehicle electric and hybrid power trains, and control of plasma fusion.

Sharing open hardware through ROP, the Robotic Open Platform

Abstract: The robot open source software community, in particular ROS, drastically boosted robotics research. However, a centralized place to exchange open hardware designs does not exist. Therefore we launched the Robotic Open Platform (ROP). A place to share and discuss open hardware designs. Among others it currently contains detailed descriptions of Willow Garage’s TurtleBot, the NimbRo-OP created by the University of Bonn and the AMIGO robot of Tech United Eindhoven. Eventually, ROP will contain a collection of affordable hardware components, allowing researchers to focus on cutting-edge research on a particular component instead of having to design the entire robot from scratch. As an example of how the Robotic Open Platform is able to facilitate this knowledge transfer, we introduce TURTLE-5k: A redesign of an existing soccer robot by a consortium of our university and companies in the wider Eindhoven area. Cooperating with industrial partners resulted in a significant cost reduction.

Frequency domain‐based nonlinearity detection and compensation in Lur'e systems

Abstract: SUMMARY Nonlinearities often lead to performance degradation in controlled dynamical systems. This paper provides a new, frequency domain-based method, for detection and optimal compensation of performance degrading nonlinear effects in Lur'e-type systems. It is shown that for such systems a sinusoidal response to a sinusoidal input is necessary and sufficient to show the existence of an equivalent linear and time invariant dynamical model that fully captures the systems’ dynamics for a well-defined set of input signals and initial conditions. This allows to quantify nonlinear effects by using a frequency domain performance measure and yields a novel method to design optimized static compensator structures that minimize performance degrading nonlinear effects. Moverover, the methods discussed in this paper allow to quantify the performance of nonlinear systems on the basis of output measurements only while requiring little knowledge about the nonlinearity and other system dynamics, which yields a useful tool to optimize performance in practice without requiring advanced nonlinear modeling or identification techniques. Finally, the theoretical results are accompanied by examples that illustrate their application in practice.Copyright © 2013 John Wiley & Sons, Ltd.

Integrating planning and execution for ROS enabled service robots using hierarchical action representations

Abstract: The aim of the RoboEarth project is to develop a globally accessible database, that enables service robots to share reusable information relevant to the execution of their daily tasks. Examples of this information are the hierarchical task descriptions, or action recipes, that represent typical household tasks as symbolic action sequences. By annotating these static action representations with hierarchical planner predicates, they can be interpreted by a Hierarchical Task Network planner such as SHOP2 to compose optimized robot plans in a flexible manner, based on the actual state of the environment and the available capabilities of the robot. To subsequently execute these plans in a household environment, the CRAM executive toolbox is adopted, allowing a tight integration between plan execution and the run-time inference of dynamically updated environment knowledge. This paper presents the integration of these two planning and execution components into one cohesive framework, tailored for the safe execution of abstract tasks in challenging household environments. The resulting framework is implemented on the AMIGO service robot and a basic experiment is conducted to demonstrate the frameworks integral functionality.

Exploiting rational basis functions in iterative learning control

Abstract: Iterative learning control approaches often suffer from poor extrapolability with respect to exogenous signals, including setpoint variations. The aim of this paper is to introduce rational basis functions in ILC. Such rational basis function have the potential to both increase performance and enhance extrapolability. The key caveat that is associated with these rational basis function lies in a significantly more complex optimization problem when compared to using polynomial basis functions. In this paper, a novel iterative optimization procedure is proposed that enables the use of rational basis functions in ILC. A simulation example confirms (1) the advantages of rational basis functions compared to pre-existing results, and (2) the efficacy of the proposed iterative algorithm.

A fast, magnetics-free flux surface estimation and q-profile reconstruction algorithm for feedback control of plasma profiles

Abstract: The flux surfaces' layout and the magnetic winding number q are important quantities for the performance and stability of tokamak plasmas. Normally, these quantities are iteratively derived by solving the plasma equilibrium for the poloidal and toroidal flux.In this work, a fast, non-iterative and magnetics-free numerical method is proposed to estimate the shape of the flux surfaces by an inward propagation of the plasma boundary shape, as can be determined for example by optical boundary reconstruction described in Hommen (2010 Rev. Sci. Instrum. 81 113504), toward the magnetic axis, as can be determined independently with the motional Stark effect (MSE) diagnostic. Flux surfaces are estimated for various plasma regimes in the ITER, JET and MAST tokamaks and are compared with results of CRONOS reconstructions and simulations, showing agreement to within 1% of the minor radius for almost all treated plasmas.The availability of the flux surface shapes combined with the pitch angles measured using MSE allow the reconstruction of the plasma q-profile, by evaluating the contour-integral over the flux surfaces of the magnetic field pitch angle. This method provides a direct and exact measure of the q-profile for arbitrary flux surface shapes, which does not rely on magnetic measurements. Results based on estimated flux surface shapes show agreement with CRONOS q-profiles of better than 10%. The impact of the shape of the flux surfaces on the q-profile, particularly the profiles of elongation and Shafranov shift, and offsets in plasma boundary and the magnetic axis are assessed.OFIT+ was conceived for real-time plasma profile control experiments and advanced tokamak operation, and provides quickly and reliably the mapping of actuators and sensors to the minor radius as well as the plasma q-profile, independent of magnetic measurements.

Optimal Energy Management for a mechanical-hybrid vehicle with cold start conditions

Abstract: This paper presents the design of an optimal Energy Management Strategy (EMS) for a hybrid vehicle that starts with a cold powertrain. The cold start negatively affects the combustion and transmission efficiency of the powertrain, caused by the higher frictional losses due to increased hydrodynamic viscosity effects. The excess fuel consumption of the engine and the excess power loss of the transmission are modeled by static relations as a function of the lubrication oil temperature. The thermodynamics in the powertrain during the heating period of the powertrain is approximated by a first-order dynamic model. The main design criterion for the optimal EMS is the minimization of the overall fuel consumption over a pre-defined driving cycle. Dynamic programming is used to find the globally optimal solution for six representative driving cycles. The results show that the cold start has a significant impact on the fuel consumption of the hybrid vehicle, yet its influence on the optimal EMS is negligible.

Modeling for Control and Optimal Design of a Power Steering Pump and an Air Conditioning Compressor Used in Heavy Duty Trucks

Abstract: The hybridization and electrification of power-trains has brought increased flexibility and, therefore, new challenges, in the design of the hybrid vehicles. Beside the main components that are used for vehicles propulsion, important energy consumers are the auxiliaries such as the power steering pump or others. The influence of these components on the fuel consumption can be defined in terms of the topology, technology and control algorithm choices. This paper presents the influence on fuel consumption of two auxiliaries for two different topologies. For this purpose models are developed for the power steering pump and for the air conditioning compressor, and validated using experimental data from components used in long-haul heavy duty trucks. A case study on the control for the power steering pump is also presented. The results show a significant fuel reduction for each component (for the power steering pump approximately 50% fuel reduction and for the ACC approximately 40% fuel reduction).

Feedforward for flexible systems with time-varying performance locations

Abstract: Increasing demands for higher throughput in high-performance motion systems, e.g. waferstages or pick-and-place machines, lead to more aggressive motion profiles (higher accelerations), a stiffer design and/or larger wafer sizes (higher mass). Therefore, larger actuation forces are required, which puts stricter demands on actuators, amplifiers and cooling. However, this design paradigm has reached the boundary of its scalability. Therefore, the next generation of high-performance motion systems are designed to be lightweight. For this class of systems the location where the tool operates, e.g. the area to be exposed or component to be placed, is constantly varying. Due to the lightweight design and the changing position of the performance output location, the system dynamics to be considered for feedforward are changing as well, i.e. the flexible modes are observed differently. This paper presents a model-based feedforward method for flexible systems with time-varying performance locations. This method is experimentally validated on a two-mass setup with flexible shaft.

High performance teleoperation using switching robust control

Abstract: The inherent transparency-stability trade-off in bilateral teleoperation poses a challenge to design controllers that find a proper balance between both requirements. Furthermore, when the environment of the teleoperation system varies within a wide range, a single controller might not be sufficient to achieve both stability and transparency. Therefore, we propose the synthesis of a switching robust controller that increases the environment stiffness range for which both stability and transparency is achieved. During the design we take into account the fact that environment estimators will have limited accuracy due to noise and uncertainty. Additionally, we use a parametric model of the human operator based on measurements of an operator's hand. We show the applicability of the approach by experiments on a 1-DOF teleoperated system interacting with a virtual spring.

Under Actuated Air Path Control of Diesel Engines for Low Emissions and High Efficiency

Abstract: This paper presents a new method for feedback control using the Exhaust Gas Recirculation (EGR) valve and Variable Geometry Turbine (VGT) of a diesel engine. The controller effectively counteracts disturbances in NOx and PM emissions while maintaining the fuel efficiency. It is shown that by using a new combination of outputs, the controlled system has very good robustness properties. Using a mean-value engine model, which is extended with an emission model, the performance of the controlled system is examined in a simulation study with various applied disturbances; the feedback controller is shown to reduce the variation of emissions and pumping losses by 80–90 %. Compared to open loop control, the feedback controlled system has lower overall emissions by 14 % in NOx, 19 % in PM and a simultaneous 0.7 % improvement of the brake specific fuel consumption is achieved.

Model predictive control of a high speed switched reluctance generator system

Abstract: This paper presents a novel voltage control strategy for the high-speed operation of a Switched Reluctance Generator. It uses a linear Model Predictive Control law based on the average system model. The controller computes the DC-link current needed to achieve the tracking of a desired voltage reference in the presence of an unknown electrical load current. Its output is converted into a pair of excitation angles (turn-on/turn-off) by means of optimized scaling and mapping, while minimizing the dominant high-speed power loss. The constraints on the average DC-link voltage and current signals are directly handled. A numerical example is provided to validate the effectiveness and the robustness of the proposed control scheme.

Model-based approach to minimize system calibrations for 3D reconstruction quality of a cone-beam CT system

Abstract: Obtaining high quality medical 3D reconstructions is of increasing importance for the medical community. The quality of a 3D reconstruction depends heavily on accurate knowledge of the position and orientation of the detector and X-ray source during a reconstruction scan. New developments in the considered class of medical imaging systems tend to lightweight design and an increasing number of scan positions. In this work, an alternative model-based approach is introduced for the currently used geometric system calibrations.

Mask side wall clamping

Abstract: Current state-of-the-art optical lithography scanners using 193nm wavelength lasers and numerical apertures of 1.35 have reached fundamental printing limits. Yet, consumer demands and device trends continue to drive smaller feature sizes, and most IC manufacturers have already navigated beyond the lithographic printing limits by turning to double patterning techniques. 1 Requiring an extra lithography step for these techniques, it is essential to keep costs down by e.g. increasing wafer throughput. Currently, leading edge immersion scanners consistently produce over 190 wafers per hour (wph). However, to keep decreasing the cost per transistor, higher throughputs of 250 wph are key targets for the year 2013 2 . Amongst others, higher throughput can be acquired by increasing acceleration of the positioning stages. One of the constraining technologies is the current mask or reticle clamping concept due to its friction based acceleration. While current reticle accelerations amount to 150 m/s 2 , some research 3 has already been performed to reticle stage accelerations of 400 m/s 2 . In this paper, a novel reticle clamping concept is presented. The concept is shown to be suitable for accelerations larger than 400 m/s 2 entirely eliminating reticle slip, whilst meeting specifications for clamping induced error with a pattern deformation of < 0.12 nm on wafer stage level (WS) and comprising high clamp stiffness.

Fast model-based control and prediction of the safety factor profile evolution in tokamak plasmas

Abstract: Advanced tokamak operation requires control over the q-profile evolution during current ramp-up and flat-top phase, as it determines both stability and performance of the plasma. This can be done by feedback controlling the q-profile evolution. For ITER, simultaneous control of the q-profile and additional control tasks needs to be demonstrated, where a governing supervisory controller (SC) needs to share the same actuators for both tasks. This implies that the q-profile controller should be able to handle real-time varying constraints on actuators and plasma physics, set by the SC. This paper describes an algorithm to control and predict the q-profile evolution using Model Predictive Control (MPC). MPC is a general optimal control technology which uses a predictive model to compute the control action and can deal routinely with real-time varying actuator and state constraints [1]. The prediction can be made available to the SC. Simulation results show the effectiveness of this approach.

Iterative learning control with rational basis functions

Abstract: Many high precision motion systems must perform repeating tasks that slightly vary. Feed forward techniques are often used to improve performance. Classical feed forward is a static calibration to compensate for known process dynamics and disturbances. Iterative learning control (ILC) is a technique which eliminates all repetitive disturbances by adapting the feed forward each repetition of the task. Classical feed forward works for any task, yet it does not take advantage of the repetitive character. Altough ILC does take this advantage, the tasks must be strictly identical, if not, transients in the feed forward may hinder practical usability. In order further to increase performance a control strategy is required that allows variations in the task and concurrently exploit the repetitive behavior.

Minimizing the number of iterations when computing a base pose for manipulation by mobile base inclusion in the inverse kinematics

Abstract: One of the most fundamental skills of a domestic service robot is the ability to manipulate objects. Hereto, the end-effector must be positioned accurately in Cartesian space. To move the end-effector to its desired pose, a suitable end-pose for the mobile platform and a corresponding configuration of the manipulator and possibly torso must be found. To determine the optimal base pose, it is common to compute manipulator configurations corresponding to a large number of possible base poses, requiring a large number of IK computations with each a number of samples or iterations with corresponding time-consuming collision checks. This paper demonstrates how the total number of required iterations can be minimized by computing a single IK solution for the kinematic chain including the base kinematics and using the result as the base pose for manipulation. As a secondary objective, the distance to joint limits is maximized. A further reduction in the total number of iterations can be achieved by only constraining translations and rotations that need to be constrained for the task at hand.