Select Page

A good understanding of the underlying mathematical methods and the results they can achieve, leads to a better comprehension of these tools.

We emphasize techniques and their usage within our optimization tools and use real life financial and industrial applications to illustrate the methods.

optimization & data science

Introduction to linear optimization

— Linear programming is an extremely powerful tool in increasingly complex economic systems in which the use of resources needs to be rationalized. Recent advances in linear programming solvers allow scientists and economists to quickly implement these techniques in a large number of operational and strategic problems. The success of such approaches depends, above all, on the choices made when modeling of the problem to be treated. This course will allow you to understand the principles behind linear optimization algorithms and to adopt the most efficient modeling approach.

Combinatorial Optimization I: Integer Programming

— The discrete nature of many decision problems can lead to a so-called combinatorial explosion. Whenever avoiding such phenomena (e.g. by relaxing integrity constraints) proves to be impossible, Integer Programming (IP) allows to tackle a great number of Combinatorial Optimization problems such as those found in the domain of logistics, production management or scheduling.

Combinatorial Optimization II: constraint programming & local search

— Whenever Integer Programming (IP) turns out to be unfit for treating a combinatorial optimization problem, it might be necessary to use the problem’s attributes in order to overcome it. Based on this concept, constraint programming and local search provide a formal framework for solving difficult combinatorial problems.

Combinatorial Optimization III: relaxation & hybridization

— Aside from the attributes, it is possible to get around a problem by using its structure. In such a case, rather than solving a large size problem subject to combinatorial explosion, it is possible to solve several small problems in a coordinated way: this is the principle of decomposition. In some cases, it may even be advantageous to combine Combinatorial Optimization techniques (IP, CP, local search) to overcome a problem particularly difficult to solve. This is the principle of hybridization.

Nonlinear optimization with Artelys Knitro: from theory to practice

— Nonlinear optimization arises in various domains such as energy, economy, finance, machine learning, model predictive control, etc. This training will enable participants to understand and practice the basics and refinements of nonlinear optimization and to model and solve problems efficiently.

Stochastic optimization and dynamic programming: applied to energy inventory management

— Inventory and financial assets management-related decisions are closely connected. One is often seeking for balance between instant profit and future gains. This course shows how dynamic programming can model similar issues.

Demand forecast models with R

— Through the sales of goods or services, Demand Forecasting is one of the key challenges in operational planning and in designing facilities for the long-term. This course provides tools to master R software used in the context of demand forecasting.

Python programming: tools for Data Science

— Data Science is the convergence of mathematics, statistics and computer science to make the most of the information contained in data. Most Artificial Intelligence (AI) methods rely on it. Python provides the Data Scientist with all the tools needed to do scientific programming. This course places particular emphasis on the quality of the code.

Introduction to Docker, Kubernetes and Serverless

— Big Data is one of the latest key issues among the challenges that face many companies. However, beyond the current trend, it is rather difficult to fully understand what Big Data is and its potential.

Software architecture, design and integration of an optimization tool

— Technical decisions related to the development, architecture, and integration of software have a strong and lasting impact on its the costs, quality, and performance. Quantitative decision support modules have specific computing features that require heavy machine resources (RAM and CPU) and complex data management. As a result, their architects require specific skills.

High performance computing and parallelization

— The simulation of complex physical systems and large-scale problem solving require massive computing power. With the introduction of Big Data, computational performance requirement increases even more. High-Performance Computing is an essential tool for research and industry. This course will focus on cluster implementation.

energy systems

Introduction to the technical and economic analysis of electrical systems and markets

— The rapid decarbonization of the electricity sector is a sine qua non condition for achieving the ambitious climate objectives that most countries have set for themselves. Quantitative analysis techniques allow for the analysis of the role different technologies can play, the revenues that players can expect according to the structure of the markets, the risks to which project developers are subject to.

Introduction to the operation of power systems

The energy transition is highlighting many upheavals in the electricity system, with the emergence of new forms of generation that are more dependent on the weather and more decentralized, as well as new forms of consumption that are more flexible and satisfy new uses, such as the electric vehicle. At the interface between production and consumption, networks are at the very core of the electrical system, and their operators are also constantly innovating to support the energy transition.

economic optimization of energy systems

Energy systems economics

The European electricity sector has undergone significant and drastic changes for several years.

These changes include evolving consumption patterns, growth of peak electricity demand, dramatic decline of the costs of renewable power plants, increasingly ambitious CO2 emission reduction targets, sector integration strategies, ambitious development plans for electrolytic hydrogen, etc. We offer a tailored training course on these evolutions in the energy sector and present the avenues proposed by national and European public institutions to tackle them: reform of support mechanisms for renewable energies, integration of markets (day ahead, capacity, CO2 quotas) at European level and strengthening of interconnections between countries.

Electricity market organization in Europe

The electricity markets address needs that correspond to specific time horizons.

Long-term via OTC, reserve and capacity markets, at the daily level for day-ahead markets and some reserve markets, on the same day for intraday markets, in real-time for balancing markets. We offer a training course that gives an overview of these different markets and an understanding of their relevance for the electricity system. We offer a focus on the way they are organized in Europe (by presenting the case of European market coupling and European balancing platforms). Practical work will be carried out in order to illustrate the economic and legal descriptions, by studying in particular what the expected valuation by a producer on each of the markets in France can be.

Energy transition and Smart Grids

The energy transition will transform the way the electricity system is operated.

The traditional situation that consists in electricity flowing from large generation groups to end-users is profoundly evolving due to decentralized production and new consumption practices (electric vehicles, load shedding, etc.). We propose a training course which focuses first on energy markets, operation of the transmission and distribution networks, principles of frequency and voltage control. A second part will then be devoted to the future challenges that can be tackled with smart grids, with particular emphasis on the case of self-consumption and the local challenges of managing flexibilities (via technological and market design solutions).

Risk management and energy systems

The key topics include forecasting and risk issues specific to the energy sector, as well as the most suitable methods for dealing with such risks.

This training course introduces the general concepts of risk management (e.g. Value-at-Risk, Stress-Testing) and its application to the specific case of energy systems, by recalling in particular certain risk hedging tools (long-term contracts, options, etc.). It also provides details on the various aspects of rigorous stochastic modeling and methodological approaches that facilitate the estimation and reduction of risk in an uncertain environment. Concrete examples from real-world problems encountered by practitioners in the energy world will facilitate the understanding and assimilation of the concepts.

Operational optimization of energy systems

This course presents the functioning of the energy market and details the different tume horizons considered by energy companies…

…when planning their production facilities (day for tomorrow, next year, 10 to 15 years ahead). The associated optimization methods, as well as the major uncertainties to be considered – electrical demand, fuel prices, weather, energy policies – will be clearly explained and illustrated by application cases.

numerical software, platforms and optimization tools

Artelys Knitro

— Artelys Knitro is a numerical software component that implements advanced nonlinear optimization techniques.

Its 4 algorithms and its numerous options allow it to offer excellent performance and great robustness in the solution of a variety of optimization problems. We offer on-demand training session that will allow you to learn how to solve nonlinear optimization problems, such as portfolio optimization, optimal network power flow, nonlinear predictive control, or Nash equilibrium models. Trusting its efficiency and robustness, hundreds of institutions worldwide have chosen Artelys Knitro to solve highly complex problems.

Artelys Kalis

— Artelys Kalis is a software component for modeling and solving large size combinatorial problems …

… through hybrid constraint programming and mathematical programming techniques. We offer on-demand training sessions that will present the principles of constraint programming and a rapid and efficient implementation of combinatorial problems of different types: tasks and timetable scheduling, resource allocation, equipment or network configuration.

FICO® Xpress Optimization Suite

— FICO® Xpress Optimization offers a complete range of modeling and numerical optimization tools.

These solutions can be quickly integrated into business problems in order to provide decision-support insights into complex problems. The following are some examples of on-demand courses that we can offer:

Logistics– Defining master plans in sectors such as transport, manufacturing, etc.
Personnel Planning– Timetabling in sectors such as aeronautics, medical, public transportation and distribution.
Networks– Defining investment strategies in sectors such as telecommunications or electricity networks, and establishing a medium-term strategy

AMPL

— AMPL is a complete and powerful algebraic modeling language for solving linear and non-linear problems with discrete or continuous variables.

We offer on-demand training sessions that will teach you how to use generic notation and familiar concepts necessary to formulate optimization problems and to examine the possible solutions. The flexibility and the ease of use of AMPL allow for a very fast prototyping and development of models, whereas its speed and options control make it a very efficient tool for repeated use in production.

FICO Xpress Insight

— FICO® Xpress Insight enables organanizations to quickly deploy any advanced analytical model as powerful applications.

Xpress Insight enables organizations to work in a collaborative environment with interactive visualizations tailored to business needs. This allows users to work with easy-to-understand models that focus on the impact of decisions on the business problem. They can share results with their peers and collaborate to make optimized decisions by performing what-if scenario analyses and comparing the impact of different strategies.

Artelys Crystal City

— Artelys Crystal City provides full support to territorial authorities in evaluating, monitoring and communicating their local multi-energy development plans.

Used today to work out the Energy Master Plans of cities such as Lyon, Grenoble, Poitiers and Pays de Gex. Artelys Crystal City provides full support to territorial authorities in evaluating, monitoring and communicating their local multi-energy development plans. At the time of the energy transition, local decision-makers are confronted with new territory planning issues where the energy dimension is a key factor in the decision-making process. We offer on-demand training sessions based on the tool Artelys Crystal City allowing to treat a variety of challenges related to topics such as energy consumption, CO2 emissions reduction, coordinating the development of distribution networks and valuating local renewable production potential.

Artelys Crystal Super Grid

— The energy sector of most countries is currently undergoing a rapid and deep mutation.

The development of renewable energies, interconnections, energy storage and demand-side response represents at the same time a challenge and an opportunity to rethink the way energy systems are operated and how we plan their evolution. Whether they are energy regulators, network operators, assets owners, researchers, all the actors have to evaluate the impacts of strategic choices that integrate this new energy reality. We offer on-demand training sessions based on Artelys Crystal Super Grid, providing quantitative elements to assess the costs and benefits of adding interconnection capacity between two countries or to optimize a national energy strategy using the investment planning module of Artelys Crystal Super Grid.

ARTELYS

DOCUMENTATION