The multiphase machines are used for various applications due to the benefits like high reliability, lower current at each phase, high fault-tolerant capacity, etc. This gives rise to the usage of multiphase machines in many fields, including electric vehicles, aircraft, and marine applications [1–3]. When the multiphase is used along with the multi-level drive system of the motor, it can be used on a high-speed system due to low voltage stress and harmonic content. The most commonly used multi-level inverters are flying capacitors, H-bridge inverters, and clamped inverters [4–6]. The open-end stator winding concept interconnected the multi-level inverter and the multiphase machines [7]. While operating under this concept, a lower switching frequency, reduced switching loss, common mode voltage elimination, and fault-tolerant capability are obtained for two-level inverters [8].
The five-phase induction motor [9] is the most promising motor structure in the research, but it has the problem of difficulty in speed control when starting. Hence, the researchers were motivated to have a six-phase induction motor design that compensates for the problem in existing multiphase machines. The optimal design of the six-phase induction motor is made by using the particle swarm optimization in the electric propulsion of submarines [10] with the squirrel cage rotor model. Also, the performance could be improved by shifting the phases of the induction motor [11] for high-power applications. The super-twisting current control scheme reduces the time delay at the coils on a discrete-time basis [12].
The speed control on a six-phase induction motor is achieved by direct torque control, sliding mode control, backstepping control, and field-oriented control [13–15]. The flux ripple and torque are reduced by controlling the inverter's duty cycle [16], which is realized using a proportional integral regulator. The voltage vectors are the major factors utilized to control the torque of a motor. Concerning the voltage vector of different switching states, a model predictive control model [17] was employed, in which two virtual voltage vectors are selected for controlled drive operation. Meanwhile, the torque and its ripple are the main factors that need to be solved for the effective operation of the induction motor so that the motor can be used in a variety of applications.
The problem focused on is reducing the complexity of handling the number of switching states of a drive system. Since the multiphase machines increase the number of switching states as lp where l is the number of levels in the drive system and p is the total phases supported by the induction motor, the need is to finely focus on each state using the controller using the subset of voltage vectors. The controller makes use of the voltage vector, while the actual action of the controller relies on the applied voltage vector. Some voltage vectors are generated, and it is not essential to use them to generate a controlled signal. This leads to the optimal selection of vector voltages used by the controller for torque ripple and harmonic reduction. Hence, an artificial neural network is used to cope with this objective here.
Similarly, the literature uses a capacitor-fed cascaded inverter [18] with a multi-stage strategy for the six-phase induction motor. But the problem is high reverse flowing current; also, handling some voltage vectors is difficult, thus increasing the complexity. Therefore, the model uses a five-level inverter, therefore, the switching states of the system are high; even using a finer and more complex scheme of voltage vector structure, the activated artificial neural system can handle these finer vectors for the optimal selection of voltage vectors. The contribution of the proposed work is listed as follows:
-
To stabilize the operation of the six-phase induction motor, a dual three-phase five-level inverter is used, which regulates the supply to the motor.
-
A novel hybrid direct control and field-oriented control is proposed to reduce torque ripples during the operation of a six-phase induction motor. This is realized by enabling both the stationary and synchronous reference frames.
-
Since the control system generates a complex and very large number of voltage vectors, an artificial neural network is implemented here for handling, which selects the switching states from the switching table.
The paper is structured as follows. Section 2 consists of a literature survey that comprises the relevant work of the six-phase induction motor. Section 3 is the design of the proposed voltage vector-based controlled switching sequence generation system concerning a five-level inverter. Section 4 is the result obtained by the proposed model compared with the existing works, and finally, the paper is concluded with section 5 where the work results are discussed.