A new Fuzzy control for two stage launch vehicle
by
Mahdy Rezaei Darestani
KNTU of technology-aerospace department
Coauthors: Mehran zareh,Hamed Hossein Afshari,Jafar Roshanian

Field of control in aerospace engineering especially in control of launch vehicles is one of the most important bases to study for developing a reliable and easy to set up control which could conduct the launch vehicle (LV) to the predetermined orbit. This action is a very heavy duty. To reach to this end a new fuzzy control have been used. One of the most important reasons for this selection is easy set up and good performance due to sever changes of plant in order to the environmental and internal variable changes. In this way first a six-degree-of-freedom simulation is developed which comes from solving the three force equations and three moment equations that for this action all the exerted body and inertial forces and moments such as aerodynamic, propulsion, gravity are calculated. Here because of launching from earth we applied the spherical rotating model of earth which leads us approximately to a real launching. For the simulation first a launch vehicle from all the available launch vehicles to specify the body characteristic such as mass, moment of inertia, dimensions, etc might be chosen. Here an LV that with the payload of approximately 2 tons could safely launch to the 300 km orbit of earth is assumed. In this way, use of fuzzy controller helped to have a stable flight with minimum deflection of controlling thrust nozzles in the entire three coupled channels of our LV.

For this purpose at each time the value of calculated angle in each channel with its predetermined values is compared which in the roll and yaw channels is equal to zero and in the pitch channel the ideal value is the pitch program.

In this fuzzy based control for determining the rule bases the error and change in error is employed which gives us 9 rule bases. Here beside of using these rules because of unacceptable control of LV to have a smoother control it's prefered to apply an online scheduling of the gains of controller in relation to the bound of error. Such that when the value of error is near the end of error bound we apply full of gain. For this purpose we used a membership function. After employing this thought the results showed that the control had been better in compared to the pure fuzzy rule bases, especially a less nozzle deflection to control of LV in all of the channels is used.

Now it is suggested in addition to this kind of fuzzy proportional gain scheduling with the aid of this method and combining with the PID controller a fuzzy PID controller is formed which its gains are determined instantaneously and finally a controller with all the characteristics of PID and Fuzzy controller is developed.

Date received: April 28, 2008

Copyright © 2008 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # caxc-92.