Vision-based control of unmanned aerial vehicles (UAVs) has been adopted in a wide range of applications due to the availability of low-cost onboard sensors and computers. Tuning such systems to work properly requires extensive domain-specific experience, which limits the growth of emerging applications. Moreover, obtaining performance limits for UAVs performing visual servoing is difficult due to the complexity of the models used. In this article, we propose a novel noise-tolerant approach for real-time identification and tuning of visual servoing systems. This is based on the deep neural networks (DNNs) classification of the system response generated by the modified relay feedback test (MRFT). The proposed method, called DNN with noise-protected MRFT (DNN-NP-MRFT), can be used with a multitude of vision sensors and estimation algorithms despite high levels of sensor noise. The response of DNN-NP-MRFT to noise perturbations is investigated and its effect on identification and tuning performance is analyzed. The proposed DNN-NP-MRFT is able to detect performance changes induced by the use of high latency vision sensors or by integrating an inertial measurement unit sensor into the UAV states estimation pipeline. Experimental identification closely matches simulation results, which can be used to explain the system behavior and predict the closed-loop performance limits for a given hardware and software setup. We also demonstrate the ability of DNN-NP-MRFT-tuned UAVs to reject external disturbances like wind or human push and pull. Finally, we discuss the advantages of the proposed DNN-NP-MRFT visual servoing design approach compared with other approaches in the literature.

This paper presents a comparison between relay feedback tuning and other controller tuning methods for a digitally controlled DC-DC buck converter operating in continuous conduction mode. In the first method, a relay feedback test is used to obtain ultimate gain and period measurements experimentally. PI and PID controllers are then tuned according to the classic Ziegler-Nichols formulae. In the second and third method, a converter model is developed based on experimental identification and analytical derivation, respectively. PI and PID controllers are then designed through optimization of an integral performance criterion applied to the simulated step response. Experimental tests show that all three approaches to controller tuning provide comparable and acceptable results, but the simplicity of the approach based on the relay feedback test makes it a viable alternative to offline-tuned optimal controllers given the ability of the former to retune the controller to adapt to system changes. Results also suggest that with some improvements to the test and tuning rules, a well-tuned controller may be produced with reasonable computational effort in the auto-tuning schema.

The chattering phenomenon associated with the use of Sliding Mode Controllers (SMC) is considered a major drawback of the robust controllers. Ideally, SMCs are to exhibit infinite switching frequency but this is always compromised by the existence of parasitic dynamics which add to the input relative degree of the system and leads to high but finite switching frequency. The performance and robustness of the controller can be closely linked to the resulting switching frequency. Often, when controlling the speed of induction motors, SMCs are applied and used to generate what is assumed to be smooth control signals. These control signals are used as modulation inputs for various carrier based modulation schemes. In doing so, the sliding action of the controller is disrupted as the SMC's switching frequency is limited by the finite carrier frequency of the modulation process. This effect is carefully analysed in this paper. An alternative is considered in this paper where SMCs are applied to the inner most control loop without the use of any carrier based modulation schemes.

Voltage source converter (VSC) plays a major role in the AC/DC interconnection. Its control stands as a significant issue as it determines the stability and performance of the system. Therefore, utilization of a control algorithm that can robustly reject disturbances is highly appreciated. This paper proposes a new approach for PI-parameters selection, used in the decoupled dq-vector control scheme, based on first-order-plus-time-delay (FOPTD) models, for which analytical formulas for optimal controller design/tuning can be obtained. Furthermore, a non-linear optimization algorithm is developed in order to achieve the optimum disturbance rejection in terms of minimizing the integral time absolute of the error (ITAE) performance index. The cost function employs a weighted-sum criterion for two sources of disturbances, namely AC input voltage fluctuation and load disturbances, which may affect the regulated voltage. Simulation results demonstrate a very similar system performance using the FOPTD tuning parameters compared to the optimum ones where both yield a robust system against disturbances. The obtained results are also verified experimentally.