Market and resource allocation algorithms with application to energy control

Sammanfattning: The energy markets of today are markets with rather few active participants. The participants are, with few exceptions, large producers and distributors. The market mechanisms that are used are constructed with this kind of a market situation in mind. With an automatic or semiautomatic approach, the market mechanism would be able to incorporate a larger number of participants. Smaller producers, and even consumers, could take an active part in the market. The gain is in more efficient markets, and – due to smaller fluctuations in demand – better resource usage from an environmental perspective.The energy markets of the Nordic countries (as well as many others) were deregulated during the last few years. The change has been radical and the situation is still rather new. We believe that the market can be made more efficient with the help of the dynamics of the small actors.The idealised world of theory (of economics) often relies on assumptions such as continuous demand and supply curves. These assumptions are useful, and they do not introduce problems in the power market situation of today, with relatively few, large, participants. When consumers and small producers are introduced on the market, the situation is different. Then it is a drawback if the market mechanims cannot handle discontinuous supply and demand.The growth in accessibility to computational power and data communications that we have experienced in the last years (and are experiencing) could be utilised when constructing mechanisms for the energy markets of tomorrow.In this thesis we suggest a new market mechanism, ConFAst, that utilises the technological progress to make it possible to incorporate a large number of active participants on the market. The mechanism does not rely on the assumptions above. The gain is a more efficient market with less fluctuations in demand over the day.To make this possible there is a need for efficient algorithms, in particular this mechanism relies on an efficient aggregation algorithm. An algorithm for aggregation of objective functions is part of this thesis. The algorithm handles maximisation with nonconcave, even noisy, objective functions. Experimental results show that the approach, in practically relevant cases, is significantly faster than the standard algorithm.