Modeling, Identification and Control Design for an Electro-Hydraulic Rotator
Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
Robotic manipulators have been introduced in industry as a form of increasing productivity. Today, there exist an interest to enlarge the application of these manipulators to outdoor environments. Forestry cranes used in the forestry industry are a clear example. A long term goal in this industry is the development of autonomous systems to increase the logging efficiency. In this thesis, we consider how to control the rotator of these cranes, which is an electro-hydraulically actuated motor, and is used to control the end ffector tool. Control system design for the rotator is a challenging task since the sensing is not available to full extent. The main reason is the harsh environment that these machines are exposed to and sensors such as encoders are very fragile and cannot be used. In this thesis we use alternative sensing devices, such as a magnetic sensor and a stereo camera. In the case of the camera we face a problem with big delay. A prediction method has been used to compute desired values. Due to various reasons certain measuring devices can not be used in the industry. We consider four cases for control system design where dfferent combinations of available sensors have been used. Initially angular position of the rotator is controlled using only the magnetic sensor. A cascade control setup is used where pressure and position are measurable, ﬁrst using the magnetic sensor and later using the camera. When only pressure measurements are available identiﬁed models have replaced sensors for position feedback. All tests and experiments have been done using a scaled version of a real forestry crane. The available crane has similar conﬁguration and dynamics as the real one and is therefor useful for experimental purposes.
Place, publisher, year, edition, pages
2010. , 60 p.
Hydraulically actuated system, Friction compensation, Cascade Control, Stereo Camera
IdentifiersURN: urn:nbn:se:liu:diva-54377ISRN: LiTH-ISY-EX--10/4216--SEOAI: oai:DiVA.org:liu-54377DiVA: diva2:303271
2010-02-08, Algoritmen, Institutionen för systemteknik Linköpings universitet 581 83 Linköping, Linköping, 10:15 (Swedish)
Shiriaev, Anton, Professor
Gunnarsson, Svante, Professor