BISC Seminar Announcement, January 29, 1998, 4pm, 310 Soda Hall

Frank Hoffmann (
Sat, 31 Jan 1998 19:22:49 +0100 (MET)

B I S C S e m i n a r A n n o u n c e m e n t


Daniel Schiffer
Technical University Berlin

Date: Thursday, January 29, 1998
Time: 4-5pm
Location : 310 Soda Hall


Co-generators are one of the most efficient and ecological energy supplies
based on fuel. In the process, a combustion engine drives a generator.
Both, the electrical power and the waste heat is used. This leads to the
best-value energy-supply for consumers who have a simultaneous need for
electrical power and heat, i.e. big industrial firms, indoor swimming
pools. However, small houses using this system face the problem that
electrical power and heat is needed at different times. They can store the
heat in the form of hot water, but because they cannot store the
electricity it needs to be fed into the in the public network. Up to now,
only very simple controls existed to switch on the co-generator when the
water in the tank is too cold and switch the unit off when it reaches a
preset upper temperature. This kind of control has huge economic

Our project results reveal that the use of fuzzy controls allows for the
most efficient operation of co-generators. Using conventional controls, it
is almost impossible to process many input data and to evaluate uncertain
data. In our project, we employed fuzzy-control because it is able to
tackle the described data. The fuzzy control measures the present need of
electrical power and heat, and estimates a future need. In combination with
other input data (i.e. inside- and outside-temperature and the information
about the payment the user will get for fed in electrical power), the
control determines an optimal economical work of the co-generator. Optimal
economical work means that the costs for running the co-generator have to
be as low as possible.

This talk presents the project's basic ideas, and describes the problem
that electrical power and heat is needed at different times. It will
explain in-depth the assembly and the function of the developed
fuzzy-control. You will also get a update of the project plans in the near

Please direct questions with regard to the contents of the talk
and request for papers to the speaker. To unsubscribe send email
to NOT !! to

Frank Hoffmann UC Berkeley
Computer Science Division Department of EECS
Email: phone: 1-510-642-8282
URL: fax: 1-510-642-5775