RotLin machines investigation on their modelling, design, control and applications

Abstract

In the future, two-degree of freedom actuators will become more popular and they can be realized in industry applications by integrating together linear or rotary motors. Generally, a linear machine and a rotary machine are employed together to realize a linear-rotary motion for manufacturing devices. Traditional integrated devices are complex and cumbersome because they use mechanical connectors or gears. Also, the positioning accuracy for linear motion and rotating movement are hard to be achieved due to the interference from the gears and connectors. That is why it is difficult to realize a highly accurate position control or angle control from simply combining two machines together. The entire volumes and masses of these devices are relatively large, and they must be followed by periodical maintenance. To improve the control accuracy in linear and rotary motion, integrated direct-drive rotary-linear (RotLin) machines have caught the eyes of scientists in recent years. Direct-drive RotLin machines can be employed in the manufacturing industry in the future for highly precise applications, such as laser cutting machines, laser soldering machines, microchip production instruments and etc. A precise simultaneous linear and rotary motion can be achieved by the RotLin machine with a compressed mechanical structure that contains both the linear and rotary parts. In this dissertation, two kinds of RotLin machines including, permanent magnet (PM) machines and a PM-free machine are investigated. Several prototypes have been produced for each type of RotLin machines. Advantages and disadvantages of these machines are compared by power density, efficiency, and cost. First, mathematics modeling of these machines is introduced according to the basic principles of the electric machine and their dynamic equations. The accuracy of the models for the machines plays a key role to control them, especially for highly precise position and angle control. The parameters of the dynamic equations directly reflect the accuracy of the modeling approach. By measuring the parameters of the machines via experiments, the mathematics model of the machines can be obtained accordingly. Also, their parameters can be calculated by the finite element method (FEM) and then modified with experience or analytical process according to their structures. Online or offline identification can be employed to get the accurate parameters of the machines. Secondly, the detail design approaches for PM RotLin machines and PM-free RotLin machines will be described. Combining magnetic circuit analysis and FEM is an effective way to design electric machines. The power output and volume, as well as specifications of the machines, can be estimated by magnetic circuit analysis at the beginning. This procedure can save a lot of time compared with FEM. In order to obtain the accurate parameters including the inductance of windings, the flux density in air gaps, the coil structures, number of turns and the force/torque outputs, etc. can be calculated and optimized through FEM. Meanwhile, the model analysis and thermal analysis can predict the performance of the machines. Therefore, the magnetic field, electric field, mechanical deformation and temperature of the machines can be accurately estimated by this method. This will help the development time for electric machines to be shortened significantly. Thirdly, a different control method was developed for the designed RotLin machines. Basically, proportional-integral-derivation (PID) control is employed for the machines. Three control loops involving the current loop, the speed loop and the position loop are built to realize a closed-loop control in governing the dynamic responses of the machines. More importantly, accurate thrust control is studied for the linear part of the machines. This is very important because industrial applications such as microchip production and smart-phone manufacturing require a high level of control from the machines. Apart from the PID control, some advanced control algorithms will be introduced and developed for the RotLin machines, including the fuzzy logic control, sliding mode control, system identification and adaptive control. The hardware simulation system dSPACE combined with the software MATLAB/Simulink will be introduced and used in developing these advanced control algorithms for the RotLin machines. Finally, sensorless control for the RotLin motors is developed. For the PM RotLin motor, the rotary part can be considered as a synchronous motor with the linear motion acting only as an observer. Principles, control scheme and simulation are all needed and tested for the sensorless control. Also, current pulse injection is also employed to calculate the linear position of the RotLin motor. Both the simulation and experimental results show the effectiveness of the developed sesnorless approach. This dissertation presents the structures and controls of RotLin motors including a rare-earth-free switched reluctance motor and permanent magnet motors. The design principles are presented. Simulation and experiments on the design and control of the motor are carried out, validated the effectiveness of the proposed RotLin motors. These motors could be widely employed in industrial equipment, robotics, electric vehicles and renewable energy generation systems in the future.