Sie sind hier:



Using the blockchain for distributed optimization

Lehrstuhl: DYN

Betreuer: L. Maxeiner (more tba),

Beginn ab: 9th October 2017

Maximale Anzahl der Teilnehmer: 5

Beschreibung: Many real world challenges can bemathematically formulated as general optimization problems. Optimizing the profits in a chemical parks, where interconnected plants are operated by different companies, is an example, with ui being the variables of the interconnecting streams.
Due to, e.g., privacy issues, where the companies want to limit the flow of information, only a distributed optimization is possible. Problems of such structure can be solved using distributed
optimization techniques such as Lagrangian decomposition, where the joint problem is solved by iterating between local optimization and a coordinated update of the exchange
variables. This procedure is well known and depicted in the figure exemplarily with local model predictive controllers.
The idea in this project is to not have a single coordinator but to do this update of the exchange variables distributedly in all systems on the blockchain using smart contracts (= code that runs on the blockchain). Thus, the necessity of a trusted coordinator is removed while the remaining properties of the hierarchical scheme are still valid.

OBJECTIVE: The objective of the project group is to use the given case study (including a
Lagrangian based distributed optimization scheme) and existing blockchain technology to
implement the coordinator of the optimization scheme as a smart contract.
TASKS: The task of this project group can be divided into the following tasks:
• Getting familiar with the blockchain technology.
• Evaluating suitable blockchain frameworks.
• Programming of the smart contracts.
• Executing the smart contracts on a test net.
• Running the smart contracts on different nodes that simulate the different agents.
• Tuining of blockchain properties.

REQUIREMENTS: We are looking for students who want to learn about, develop skills in, and work with the blockchain technology. All participants should have programming skills
and used Linux based operating systems before.