The following information was submitted:
Transactions: INTERNATIONAL JOURNAL of CIRCUITS, SYSTEMS and SIGNAL PROCESSING
Transactions ID Number: 19-463
Full Name: Mihai Lungu
Position: Lecturer
Age: ON
Sex: Male
Address: Carol Blv., no. 6, Craiova, Dolj, Romania
Country: ROMANIA
Tel:
Tel prefix:
Fax:
E-mail address: Lma1312@yahoo.com
Other E-mails: mlungu@elth.ucv.ro
Title of the Paper: The stabilization and the identification of the rockets' movement in vertical plane
Authors as they appear in the Paper: Mihai Lungu
Email addresses of all the authors: Lma1312@yahoo.com,mlungu@elth.ucv.ro
Number of paper pages: 10
Abstract: The paper presents some angular stabilization systems of the rockets in vertical pane using differential or integrator gyroscope. The first system has not a correction subsystem, while the second one has. One has determined the transfer functions (in closed loop or in open loop) of the two systems. The positioning of the systems' eigenvalues proofs the systems' stability. The systems respond very fast to a step input – the duration of the transient regime, for the two systems, is about one second. Using three different methods (least square method, instrumental variables' method - MVI and neural networks method), one makes the identification of the system. For both systems one obtains, using a Matlab/Simulink program, the frequency characteristics, indicial functions in the complex plane and in discrete plane, responses to impulse input in the complex and discrete planes. With least square method (LSM) the output of the system and the output of the model for the tw!
o systems were plotted. The identification is made very well – the two signals overlap. With the second identification method, one obtained the frequency characteristics for LSM and MVI on the same graphic. The identification is made using neural networks. Using this method, one obtained the indicial responses of the systems and of the neural networks (these signals overlap too), the weights and the biases of the neural networks and so on. The system's identification made also be done using the prediction error method (MEP). This method is more complicated than the others, but it is more precisely. The author also presents other two systems for rockets' stabilization: systems with accelerometer and correction subsystem (figures 16 and 17). These two systems also give good stabilization results.
Keywords: rockets' movement, stabilization, identification methods, differentiator gyroscope, neural network
EXTENSION of the file: .doc
Special (Invited) Session: Angular stabilization systems of the rockets in vertical plane using differentiator gyroscope
Organizer of the Session: 201-113
How Did you learn about congress: Romulus Lungu romulus_lungu@yahoo.com
IP ADDRESS: 193.231.39.130
