• P. Sopasakis, K. Menounou and P Patrinos, SuperSCS: fast and accurate large-scale conic optimization, IEEE European Control Conference (ECC), 2019 (software).
• P. Sopasakis, M. Schuurmans and P. Patrinos, Risk-averse risk-constrained optimal control, IEEE European Control Conference (ECC), 2019 (software, documentation).
• P. Sopasakis, A. Themelis, J. Suykens and P. Patrinos, A Primal-dual line search method and applications in image processing, European Signal Processing Conference 2017 (arxiv, bibtex)
• L. Stella, A. Themelis, P. Sopasakis and P. Patrinos, A simple and efficient algorithm for nonlinear model predictive control, IEEE CDC, 2017.
• A.K. Sampathirao, P. Sopasakis, A. Bemporad and P. Patrinos, Proximal limited-memory quasi-Newton methods for scenario-based stochastic optimal control, IFAC 2017 (arxiv, bibtex).
• A.K. Sampathirao, P. Sopasakis, A. Bemporad and P. Patrinos, GPU-accelerated stochastic predictive control of drinking water networks, IEEE Trans. Control Systems Technology 26(2):551-562, 2018 (arxiv, bibtex).
• A.K. Sampathirao, P. Sopasakis, A. Bemporad and P. Patrinos, Fast parallelizable scenario-based stochastic optimization, 4th European Conference on Computational Optimization, Leuven, Belgium, Sept. 2016 (slides, bibtex).
• A. K. Sampathirao, P. Sopasakis, A. Bemporad and P. Patrinos (2015), Distributed solution of stochastic optimal control problems on GPUs, 54 IEEE Conf. Decision and Control, Osaka, Japan, Dec 2015 (pdf, slides, errata, bibtex).
• P. Patrinos, P. Sopasakis and H. Sarimveis, A global piecewise smooth Newton method for fast large-scale model predictive control, Automatica 47(9), pp. 2016-2022. (bibtex)
• P. Sopasakis, N.Freris and P. Patrinos, Accelerated reconstruction of a compressively sampled data stream, 24th European Signal Processing conference, Budapest, Hungary, 2016 (pdf, slides, bibtex).

SuperSCS: Fast and accurate conic solver

SuperSCS is a fast solver for large-scale conic problems of the form

\begin{eqnarray*} && \mathrm{Minimize}\ c' x \\ &&Ax + s = b\\ &&s\in\mathcal{K}, \end{eqnarray*}

where \(A\in\mathbb{R}^{m\times n}\) is a (sparse) matrix, and \(\mathcal{K}\) is a closed, convex, proper cone. SuperSCS is based on the algorithmic scheme SuperMann applied to a Douglas-Rachford splitting on the self-dual homogeneous embedding of the original problem.

SuperSCS achieves higher accuracy and faster convergence compared to SCS as you can see for example here.

Risk-averse Model Predictive Control

Risk-averse model predictive control (MPC) offers a control framework that allows one to account for ambiguity in the knowledge of the underlying probability distribution and unifies stochastic and worst-case MPC.

Risk-averse MPC can be employed in presence of ambiguity on the knowledge of the actual probability distribution of the system disturbances.

We have studied risk-averse MPC problems for constrained nonlinear Markovian switching systems using generic cost functions, and derive Lyapunov-type risk-averse stability conditions by leveraging the properties of risk-averse dynamic programming operators.

We propose a controller design procedure to design risk-averse stabilizing terminal conditions for constrained nonlinear Markovian switching systems.

Additionally, we cast the resulting risk-averse optimal control problem in a favourable form which can be solved efficiently and thus deems risk-averse MPC suitable for embedded applications.

NEW: Checkout our new MATLAB toolbox that allows the formulation and solution of general risk-averse problems with risk constraints.

RapidNet

RapidNet is an open-source software written in CUDA-C++ involving open-source libraries (cuBLAS and rapidjson) which allows the easy design of stochastic model predictive controllers for drinking water networks taking into account the volatility associated with water demands and price uncertainty in a deregulated bid-based energy market.

Some of the features of RapidNet are:

1. Efficient operating management of drinking water networks
2. Uncertainty on demands and electricity prices
3. Mitigation of the volatility in electricity cost
4. Very fast Stochastic Model Predictive Control
5. Parallelization on NVIDIA Graphics Processing Units (GPU)

Embedded Nonlinear MPC

This is an ANSI C embedded control framework for nonlinear model predictive control which implements a very fast nonconvex optimization solver with convergence guarantees that significantly outperforms SQP/SCP approaches. It is particularly suitable for robotic applications.

In this video, the embedded nonlinear predictive controller is used to drive an autonomous ground vehicle...

and on a micro aerial vehicle...

Fast Online Compressive Sensing

Compressed sensing is a signal processing methodology for the reconstruction of sparsely sampled signals and it offers a new paradigm for sampling signals based on their innovation, that is, the minimum number of coefficients sufficient to accurately represent it in an appropriately selected basis. Compressed sensing leads to a lower sampling rate compared to theories using some fixed basis and has many applications in image processing, medical imaging and MRI, photography, holography, facial recognition, radio astronomy, radar technology and more.

The traditional compressed sensing approach is naturally offline, in that it amounts to sparsely sampling and reconstructing a given dataset. Recently, an online algorithm for performing compressed sensing on streaming data was proposed (Freris et al., 2013): the scheme uses recursive sampling of the input stream and recursive decompression to accurately estimate stream entries from the acquired noisy measurements.

In this work we developed a recursive compressed sensing algorithm which makes use of a Forward-Backward Newton (FBN) method to efficient solve the involved LASSO problems in real time. Our simulations showed significant speed-up compared to all well-established algorithms for such problems such as (F)ISTA, ADMM and the L1LS algorithm by Kim et al. Exploiting the fact that in recursive LASSO the optimizer at time k can be used to produce a good estimate at time k+1, and using the fact the FBN converges to the solution locally quadratically, at every sampling time we only need to perform a few Newton iterations.

Jaqpot Quattro

JAQPOT Quattro is a comprehensive open-source web application for linking molecular structure with biological properties and effects of chemical compounds and nanomaterials to living organisms. JAQPOT Quattro uses machine learning algorithms to extract knowledge out of structured databased.

Stochastic control of large-scale water networks

Drinking water networks are large-scale systems whose operation is characterised by complex uncertainty patterns. Pressures across the network need to be kept in certain limits, tanks must not overflow and water demand requirements must be met, while unexpectedly high demands and electricity prices may signal the alarm in the control centre of the network. Water demand and energy price are fluctuating uncertain parameters that deem the control of these systems rather challenging and dictate computationally intense operations for the determination of how valves and pumps need to be operated for a smooth, economic and safe operation of the whole infrastructure.

We have proposed as optimisation-based control scheme which accounts for the aforementioned uncertainty. The resulting optimisation problems involve more than 1 million decision variables and has to be solved in real time. Modern high-end CPUs require up to several minutes for their solution; We employed an accelerated proximal gradient (APG) algorithm which can be vastly parallelised and solved on a GPU to afford computation times down to a few tens of a second! These developments pave the way for the use of optimisation-based control algorithms (such as stochastic model predictive control) in large-scale water networks which - as simulations show - can lead to a significant reduction of energy consumption, higher quality of service and a smoother operation of the network.