Control of a Linear Tubular Permanent Magnet Motor

Abstract

In this thesis a linear permanent magnet motor has been examined. The motor is designed by Resonator AS is a prototype. The motor is designed for usage in offshore drilling where maintenance demands are high. This linear motor has no need for mechanical parts like bearings, valves and such which are prone for high maintenance cost, making the motor more suitable. In previous work the motor was modeled mathematically. In this thesis the model was first tested against a prototype in the lab, then the model was verified and then two different control schemes were established and simulations were done to test them.The linear model that was tested against the real motor proved to be too simplified. The linearization of the gas springs was to crude and it is suggested that the model should use the non-linear equations for gas springs instead. Also the model of the impact proved to be wrong. Another model is suggested, but not tested. This uses energy conservations laws instead. Two different control objectives and strategies were established. Both control systems uses emf created by the piston as feedback. This ensures that the motor runs around its natural frequency. The goal for both of them is to let the motor run mostly at a sinusoidal motion and get the highest possible impact speed. Both controllers are using a PI-regulator. Unfortunately none of the controllers could be tested in the lab because of a damaged position sensor.The first controller tries to eliminate a double impact that happens each period. Using a defined time called collision time; the controller will minimize the difference between the collision time and a reference time. When the collision time is higher than the reference, the amplitude of the input voltage is increased. And if it is lower than the reference it is decreased. The controller is able to eliminate the second impact and the motor runs more smoothly however the speed at which the collision occurs is not the maximum speed of the stator. The objective of the second controller is to make sure the collision happens when the stator speed is at its maximum. To ensure this the difference between collision time and the time at which maximum speed of the stator occurs is measured. Then the differenceerror is minimized using a regulator. The second controller shows significant increase in impact speed and thus an increase in energy transferred to the rock making it more suitable for drilling. The second control system is more suitable for the intended application. Implementing the control system could prove difficult because they must measure the time of collision.