A5057375078- Advanced Control Systems Optimization
- Advanced Optimization Algorithms Research
- Fault Detection and Control Systems
- Control Systems and Identification
- Aerospace Engineering and Energy Systems
- Spacecraft Dynamics and Control
- Process Optimization and Integration
- Iterative Learning Control Systems
- Real-time simulation and control systems
- Numerical methods for differential equations
- Robotic Path Planning Algorithms
- Stability and Control of Uncertain Systems
- Optimization and Variational Analysis
- Sparse and Compressive Sensing Techniques
- Robotic Mechanisms and Dynamics
- Probabilistic and Robust Engineering Design
- Adaptive Control of Nonlinear Systems
- Vehicle Dynamics and Control Systems
- Stochastic Gradient Optimization Techniques
- Computational Fluid Dynamics and Aerodynamics
- Advanced Multi-Objective Optimization Algorithms
- Advanced Aircraft Design and Technologies
- Aerospace Engineering and Control Systems
- Microbial Metabolic Engineering and Bioproduction
- Fuel Cells and Related Materials
University of Freiburg
2016-2025
Microsystems (United Kingdom)
2015-2024
HTWG Hochschule Konstanz - Technik, Wirtschaft und Gestaltung
2024
Mitsubishi Electric (United States)
2024
Fraunhofer Institute for Solar Energy Systems
2022
Universidad Carlos III de Madrid
2021
University of Fribourg
2021
Walter de Gruyter (Germany)
2020
Google (United States)
2020
Japan Science and Technology Agency
2020
Abstract In this paper the software environment and algorithm collection ACADO Toolkit is presented, which implements tools for automatic control dynamic optimization. It provides a general framework using great variety of algorithms direct optimal control, including model predictive as well state parameter estimation. The implemented self‐contained C++ code, while object‐oriented design allows convenient coupling existing optimization packages extending it with user‐written routines. We...
An efficient Newton-type scheme for the approximate on-line solution of optimization problems as they occur in optimal feedback control is presented. The allows a fast reaction to disturbances by delivering approximations exact which are iteratively refined during runtime controlled process. contractivity this real-time iteration proven, and bound on loss optimality---compared with theoretical solution---is given. robustness excellent performance method demonstrated numerical experiment, an...
Standard model predictive control (MPC) yields an asymptotically stable steady-state solution using the following procedure. Given a dynamic model, steady state of interest is selected, stage cost defined that measures deviation from this selected state, controller function summation over time horizon, and optimal shown to be Lyapunov for closed-loop system. In technical note, arbitrary economic objective, which may not depend on For class nonlinear systems costs, note constructs suitable...
This paper focuses on time-optimal path tracking, a subproblem in motion planning of robot systems. Through nonlinear change variables, the tracking problem is transformed here into convex optimal control with single state. Various convexity-preserving extension are introduced, resulting versatile approach for tracking. A direct transcription method presented that reduces finding globally trajectory to solving second-order cone program using robust numerical algorithms freely available....
The goal of this paper is to demonstrate the capacity model predictive control (MPC) generate stable walking motions without use predefined footsteps. Building up on well-known MPC schemes for motion generation, we show that a minimal modification these allows designing an online generator can track given reference speed robot and decide automatically footstep placement. Simulation results are proposed HRP-2 humanoid robot, showing significant improvement over previous approaches.
Global high-precision atmospheric Δ14CO2 records covering the last two decades are presented, and evaluated in terms of changing (radio)carbon sources sinks, using coarse-grid carbon cycle model GRACE. Dedicated simulations global trends interhemispheric differences with respect to CO2 as well δ13CO2 Δ14CO2, shown be good agreement available observations (1940–2008). While until 1990s decreasing trend was governed by equilibration bomb 14C perturbation oceans terrestrial biosphere, largest...
Linear model predictive control (MPC) can be currently deployed at outstanding speeds, thanks to recent progress in algorithms for solving online the underlying structured quadratic programs. In contrast, nonlinear MPC (NMPC) requires deployment of more elaborate algorithms, which require longer computation times than linear MPC. Nonetheless, computational speeds NMPC comparable those are now regularly reported, provided that adequate used. this paper, we aim clarifying similarities and...
This paper is about distributed derivative-based algorithms for solving optimization problems with a separable (potentially nonconvex) objective function and coupled affine constraints. A parallelizable method proposed that combines ideas from the fields of sequential quadratic programming augmented Lagrangian algorithms. The negotiates shared dual variables may be interpreted as prices, concept employed in decomposition methods alternating direction multipliers (ADMM). Here, each agent...
This paper introduces HPIPM, a high-performance framework for quadratic programming (QP), designed to provide building blocks efficiently and reliably solve model predictive control problems. HPIPM currently supports three QP types, provides interior point method (IPM) solvers as well (partial) condensing routines. In particular, the IPM optimal QPs is intended supersede HPMPC solver, it largely improves robustness while keeping focus on speed. Numerical experiments show that solves...
A Newton-type method is investigated for online optimisation in nonlinear model predictive control, the so-called real-time iteration scheme. Only one performed per sampling instant this scheme, and control of system solution optimal problem are parallel. In resulting combined dynamics optimiser, actual feedback each step based on current estimate, estimates at refined transferred to next by a specially designed transition. This approach yields an efficient algorithm that has already been...
A novel optimization method is proposed to minimize a convex function subject bilinear matrix inequality (BMI) constraints. The key idea decompose the mapping as difference between two positive semidefinite mappings. At each iteration of algorithm concave part linearized, leading subproblem. Applications various output feedback controller synthesis problems are presented. In these applications, subproblem in step can be turned into problem with linear (LMI) performance has been benchmarked...
We address the problem of real-time obstacle avoidance on low-friction road surfaces using spatial Nonlinear Model Predictive Control (NMPC). use a nonlinear four-wheel vehicle dynamics model that includes load transfer. To overcome computational difficulties we propose to ACADO Code Generation tool which generates NMPC algorithms based iteration scheme for dynamic optimization. The exported plain C code is tailored dynamics, resulting in faster run-times effort feasibility. advantages...
Building on previous propositions to generate walking gaits online through the use of linear model predictive control, goal this paper is show that it possible allow top a continuous adaptation positions foot steps, allowing generation stable even in presence strong perturbations, and additional requires only minimal modification schemes, especially maintaining same form. Simulation results are proposed then HRP-2 humanoid robot, showing significant improvement over schemes.