email: jhonny.gonzalez at manchester . ac . uk
I am currently a postdoctoral Research Associate at The University of Manchester working under the supervision of Goran Peskir and John Moriarty.
I did my PhD in Mathematics within John Moriarty's energy group at The University of Manchester. My PhD was funded by the EPSRC project The Autonomic Power System.
I also hold a MSc in Mathematical Finance (also from Manchester) and a BSc in Mathematics (Analysis and Topology) from the National University of Colombia in Bogota.
On my spare time I enjoy playing the piano (10+ years) and the accordion piano. Yann Tiersen can be blamed for my keenness on minimalist music!
My research interests include stochastic control (under risk-neutral and risk-averse objectives), optimal switching, optimal stopping, Markov Chain Monte Carlo methods, stochastic modelling of energy markets and machine learning. I am particularly focused on applications to industrial problems in the energy sector.
During my postdoc I worked on the Upside project,
developing statistical models for CO2 emissions associated with electricity generation in the UK,
as well as stochastic control algorithms for optimising energy storage.
This effort was an exciting collaboration with Upside Energy
and three other industrial partners.
My current focus is on the development of new techniques (based on optimal stopping and Neural Networks) for solving quickest detection and sequential testing problems in
the context of transient stability for power systems.
During my PhD I conducted research on the risk management of energy assets operating in liberalised markets. I developed a new risk-sensitive stochastic control methodology, extending risk-neutral optimal switching problems with entropic risk measures, and apply it to a range of examples including energy storage, district energy systems (CHP units + boilers + storage) and power plants. I also developed MCMC algorithms for estimating energy spot price models based on sum of Ornstein-Uhlenbeck processes. The algorithms were applied to the UK electricity and gas markets and to the German EEX electricity market.
My scientific computing toolkit includes Matlab, Python and C++. I have experience with GPU/Cuda programming, and the integration of research code into a production environment. For non technical computing I have used Java, C++ (with the Qt framework), Python, PyQt, and Javascript. My github account lists some of my open source code.