الفهرس 
COVEROnly 14 pages are availabe for public view
 |  | from | 127 |  |  |
| | | from | 127 | | |
Abstract
The transverse flux linear motor (TFLM) drive is nowadays one of the most interesting electric drives due to its good features that made it comparable with the other electric drives, for certain application.
In this thesis, The dynamic response of an TFLM has been obtained using information derived from the finite element analysis (FEA), which can determine the exciting coil flux-linkages as a function of both the coil current and the linear position of the mover. This work has been achieved through a digital simulation study of the mathematical model of the motor.
This model comprises a set of phase circuit equations in addition to the mechanical differential equations of motion. These equations are then solved using numerical integration of the nonlinear differential equations of the motor using the magnetization data expressed in the form of a look-up table Ψ(x, i). The cubic spline interpolation is used to determine the intermediate values of the variables given by the look-up table, which, in turn, gives more accurate representation than other methods. This method is applied to compute the instantaneous values of the current and force for each phase and the total thrust force.
Early “traditional” controllers for TFLM have a simple control technique. This control technique causes ripples in the thrust force and speed profiles to take place. Since the thrust force developed by a TFLM is a nonlinear function of phase currents and mover position such ripples often occur. These ripples represent one of the main disadvantages of this type of motors.
The main contribution of this thesis is the development of an artificial neural network controller (ANN) for transverse flux linear motors for the purpose of speed control. The dynamic response of the TFLM with the proposed controller is studied during starting, and under different load disturbances. The effectiveness of the proposed artificial neural network controller is then compared to both of the conventional proportional plus integral plus derivative controller (PID), and the fine tuning of the PID controllers by using Ziegler-Nichols Rules. The dynamic response of the TFLM with this proposed controller is found to be fast and of high speed response.