Advanced control of electrical machines in marine applications

• Main Author: Zhai, Jianyang
• Other Authors: Wang Youyi
• Format: Theses and Dissertations
• Language: English
• Published: 2019
• Subjects: Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/85927; http://hdl.handle.net/10220/50456
• Institution: Nanyang Technological University

With the rapid development of power electronic devices, the ideas of more-electric ships (MES) and all-electric ships (AES) have become practical. A shipboard power system (SPS) permits more efficient utilization of the energy provided by fossil fuel. Electricity can be easily transported through cables or wires, which considerably saves space and reduces weight. This update creates an impact on almost every part of the system, including the power distribution system, the propulsion system, and the auxiliary system. Electric machines (both thrusters and deck machinery products) are the most critical parts on an SPS. The traditional direct connection between the prime mover and the propeller is replaced with an electrical propulsion system. The operation speed for an electrical propulsion system is not fixed. Thus it can run at any optimized speed to reduce fuel consumption. The dynamic performance of an electric machine is better than diesel engines or gas turbines. The whole system becomes more robust, and the maintenance cost is minimized. These advantages would not have been possible without the development of electric machine drives. This thesis is focused on improving the performance of electric machines. To begin with, in response to the complicated environment in marine applications, requirements to the dynamic performance are higher. A novel algorithm named the optimal reset controller (ORC) is introduced to the speed controller of electric machines. The ORC continuously resets the state of the base system through a designed optimal law. Through which, the transient response is improved while the overshoot is eliminated. The algorithm is shown in detail. It is verified by hardware-in-the-loop (HIL) tests and experiments conducted on a 15 kW induction machine test rig. Next, regarding the extreme operation conditions in marine applications, the demand for the permanent magnet synchronous motor (PMSM) sensorless control is growing. Since the traditional back-EMF sensorless control method has limited performance in low speed, a common approach is to start the machine in an open loop and switch to closed loop sensorless controller when the machine is accelerated. The switching transient may cause current overshoot and machine vibration. The switching transient is comprehensively analyzed, and a smooth transition method is proposed. It is a closed loop approach to reduce the q-axis current gradually with the error angle as the feedback. The simulation verifications are shown. Hardware experiments are conducted on two sets of PMSM test rigs rated 3 kW and 50 kW, respectively. In marine applications, parameter variations have always limited the system performance and jeopardizing the robustness. Electric ships are required to work in diverse environments. They may have to work under a considerable variation of temperatures. In addition, they may endure a long-term operation without maintenance during ocean-going voyages. On the induction machine (IM) part, the parameter variation conditions are reviewed. The rotor resistance $R_r$ is the most important parameter in the slip speed calculation of indirect field oriented control (IFOC). The DQ misalignment that caused by parameter variations is analyzed, a novel method is proposed to estimate the DQ misalignment angle, and compensate it through an updated $R_r$. In the simulation, the influence of parameter change on the DQ misalignment angle is shown, and the effectiveness of the proposed method is verified. Hardware tests are carried out on the 15 kW IM. Error angles under different rotor time constants are calculated to test the proposed method. Equivalently, the effect of parameter variation on a PMSM is discussed, and the on-line identification methods are presented. A rotor flux Luenberger observer is introduced. The equations are derived, and the hardware test results are reviewed. There is a 6\% error between the estimation and the measurement. The possible causes are discussed. Next, an on-line recursive least square (RLS) method is applied to identify multiple PMSM parameters at the same time. Experiment results are shown. It is proved that this method is highly sensitive to the test environments. The effect of the d-axis current perturbation and the setting of the forgotten factor are very important parts for the algorithm design. They are discussed thoroughly in this chapter. Simulation and experiment results are conducted on the 3 kW PMSM test rig.