**4. Discussion**

In the present study, the deterministic signal was used as the reference signal and the stochastic signal was used as the interference.

This circumstance can be explained by the following two factors:


The second factor also explains the choice of only the current control channel of the vector control system for the investigation.

The quality of the current control channel of the vector control system for the asynchronous traction drive was determined according to its transient characteristics under the condition of absence of interferences at the input (Figure 7a). Comparison of the quality indicators of the basic system and of the system with the optimal controller (Table 3) showed that:


The accuracy of the overshooting of the current control channels was determined based on the transient functions of the systems in the presence of 'white-noise'-type interference at the controller input (Figure 7b). The comparison of simulation results (Table 3) showed that the system with the optimal controller had a 1.3% lower control error than the basic system. This fact suggests that the power loss caused by control interferences is 1.69% lower in the traction drive with the optimal controller than in the drive with the basic control system. This can be supported by the phase current diagrams of the stator (Figure 6) under the presence of interferences at the controller input. For the steady-state mode, the phase current value of the stator of the traction motor with the basic control system was 444 A, with the optimal controller—450 A.

There are some important caveats to the study that deserve mention:


Nonetheless, it should be taken into account that the models of the current control channels were based on the assumption that the reference signal at the current controller input was a deterministic signal. This factor implies certain limitations to the application of the model developed. To account for this factor, additional investigations must be conducted. At the same time, the authors realize the difficulties related to retrieval of the experimental data under the conditions of operation of an electric locomotive.

In the process of work on the present paper, the authors encountered objective difficulties related to the absence of any possibility to conduct a full-scale experiment and retrieve valid experimental data. This is related to the fact that the control systems are essentially microprocessor systems by design. To record the experimental oscillograms of the signals at the current control input, it is necessary to obtain permission from the respective manufacturers of the systems. This is difficult to implement as this is related with trade secrets.

Further work to improve on these new developments is suggested:

