print page

Marcus Venzke

Picture of Marcus Venzke
Dr. Marcus Venzke
Room 4.086, building E
Am Schwarzenberg-Campus 3
21073 Hamburg
phone+49 40 42878 - 3378
fax+49 40 427 - 3 - 10456
e-mail

I am in the Institute of Telematics since 1997, having the position of the senior engineer (Oberingenieur) since 2004. Find my private homepage at www.MarcusVenzke.de.

Teaching

Publications

Marcus Venzke, Daniel Klisch, Philipp Kubik, Asad Ali, Jesper Dell Missier and Volker Turau. Artificial Neural Networks for Sensor Data Classification on Small Embedded Systems. Technical Report Report arXiv:2012.08403, arXiv.org e-Print Archive - Computing Research Repository (CoRR), Cornell University, December 2020.
[BibTeX] [Abstract] [URL] [PDF]
@TechReport{Telematik_Venzke_ANNsES, author = {Marcus Venzke and Daniel Klisch and Philipp Kubik and Asad Ali and Jesper Dell Missier and Volker Turau}, title = {Artificial Neural Networks for Sensor Data Classification on Small Embedded Systems}, number = {Report arXiv:2012.08403}, institution = {arXiv.org e-Print Archive - Computing Research Repository (CoRR)}, address = {Cornell University}, month = dec, year = 2020, }
Abstract: In this paper we investigate the usage of machine learning for interpreting measured sensor values in sensor modules. In particular we analyze the potential of artificial neural networks (ANNs) on low-cost microcontrollers with a few kilobytes of memory to semantically enrich data captured by sensors. The focus is on classifying temporal data series with a high level of reliability. Design and implementation of ANNs are analyzed considering Feed Forward Neural Networks (FFNNs) and Recurrent Neural Networks (RNNs). We validate the developed ANNs in a case study of optical hand gesture recognition on an 8-bit microcontroller. The best reliability was found for an FFNN with two layers and 1493 parameters requiring an execution time of 36 ms. We propose a workflow to develop ANNs for embedded devices.
Tobias Lübkert, Marcus Venzke and Volker Turau. Calculating retail prices from demand response target schedules to operate domestic electric water heaters. Energy Informatics, 1(1):31, October 2018.
[BibTeX] [Abstract] [DOI]
@Article{Telematik_EI_2018, author = {Tobias L{\"u}bkert and Marcus Venzke and Volker Turau}, title = {Calculating retail prices from demand response target schedules to operate domestic electric water heaters}, pages = 31, journal = {Energy Informatics}, volume = {1}, number = {1}, day = {10}, month = oct, year = 2018, }
Abstract: The paper proposes a demand response scheme controlling many domestic electric water heaters (DEWHs) with a price function to consume electric power according to a target schedule. It discusses at length the design of an algorithm to calculate the price function from a target schedule. The price function is used by the control of each DEWH to automatically and optimally minimize its local heating costs. It is demonstrated that the resulting total power consumption approximates the target schedule. The algorithm was successfully validated by simulation with a realistic set of 50 DEWHs assuming perfect knowledge of parameters and water consumption. It is shown that the algorithm is also applicable to clusters of large numbers of DEWHs with statistical knowledge only. However, this leads to a slightly higher deviation from the target schedule.
Tobias Lübkert, Marcus Venzke, Nhat-Vinh Vo and Volker Turau. Understanding Price Functions to Control Domestic Electric Water Heaters for Demand Response. Computer Science - Research and Development, 81–92, February 2018.
[BibTeX] [Abstract] [DOI]
@Article{Telematik_Demand_Response_DEWH_2017, author = {Tobias L{\"u}bkert and Marcus Venzke and Nhat-Vinh Vo and Volker Turau}, title = {Understanding Price Functions to Control Domestic Electric Water Heaters for Demand Response}, pages = {81-92}, journal = {Computer Science - Research and Development}, volume = {}, month = feb, year = 2018, }
Abstract: A well-known mechanism for demand response is sending price signals to customers a day ahead. Customers then postpone or advance their usage of electricity to minimize cost. Setting up price functions that adapt the customers' load to availability is a big challenge. This paper investigates the feasibility of finding day-ahead price functions to induce a desired load profile of Domestic Electric Water Heaters (DEWHs) minimizing their electricity cost for demand response. Bilevel optimization is applied for a single DEWH using a simplified linear model and full knowledge. This leads to a solvable bilevel problem and allows understanding optimality of price functions and resulting heating profiles. It is shown that with the resulting price functions the DEWH may select many significantly different heating profiles leading to the same cost. Thus the price does not uniquely induce the desired heating profile. The acquired knowledge forms the basis for a procedure to create price functions for controlling the load profile of many DEWHs.

The complete list of publications is available separately.

Supervised Theses

Ongoing Theses

Completed Theses