فهرست مطالب a. banazadeh

In this study, in order to enhance the accuracy of tracking repetitive maneuvers in Unmanned Aerial Vehicles (UAVs), a learning-based control scheme is proposed. At the outset, the controller is designed based on the sliding mode control (SMC) technique. In addition, the offline PD-type memory-based iterative learning control (ILC) is used along with SMC. The purpose of using ILC method is to reduce the effect of system uncertainty on the controller and decrease repetitive errors by adjusting the input control signal to dynamics and thus, to increase the reliability of following the desired path. In the ILC scheme, the error of states is saved during the maneuvers which will be used in the subsequent iteration. Also, in order to increase flexibility of the new control structure, ILC-SMC, a multilayer perceptron (MLP) has been developed. This network is designed to extend the control signal, generated by ILC, to similar maneuvers. The inputs of this neural network are the initial conditions for starting the maneuver and the output of the neural network is a gain that is multiplied by the stored control signal ILC and produces a new control signal. This generated signal will be suitable for similar maneuvers. The Levenberg–Marquardt (LM) algorithm has been used to train the multilayer perceptron artificial neural network. This method was then used in loop maneuvers. In this simulation, the difference between the maneuvers was in the acceleration of the maneuver, the radius of the maneuver, and the initial speed of the maneuver. This reduced the tracking error for similar maneuvers without performing the training process for the ILC control component. The presented control scheme is applied to a quadrotor aerial vehicle for tracking desired trajectories and it is shown that the vehicle is able to follow the desired trajectory better than the conventional SMC in the presence of uncertainties.

One of the reasons for the increase in induced drag is the vortices created at the wing tip, which has a significant effect on reducing aerodynamic efficiency. Therefore, in order to reduce vortices and the induced drag as well as to improve the aerodynamic performance, the use of wing grid is recommended. Wing grids perform better at low Reynolds numbers, and combination of parameters such as taper ratio, aspect ratio, and twist has a better effect on wing performance and reducing turbulence intensity and induced drag. The purpose of this paper is to improve the aerodynamic performance of compound wing using the wing grids. In this study, the numerical and experimental approaches have been used to investigate the effect of these parameters and also, two key parameters: the grid dihedral angle and sweep angle. Also, a force balance test has been performed for force analysis and numerical solution validation. Wing grid dihedral angle decreases induced drag by increasing the space between separated tip vortices and prevents reinforcing effects due to superposition. On the other hand, dihedral angle should be arranged to increase the aerodynamic efficiency. In other words, increase in dihedral angle may defect the overall performance of the wing. The optimum configuration is found to be symmetric, where the dihedral distribution with a 40° angle for the first grid is reduced gradually to a value of -20° for the last one. In addition, sweep angle distribution for the obtained optimized dihedral angle is also investigated. Initially, each grid span is decreased from the first grid to the last at a constant rate. This increases the sweep angle and enhances the aerodynamic efficiency by 15%. Furthermore, the span of the side grids is reduced from the middle grid and marching the wing leading and trailing edges. Elliptical wing configuration has also been shown to increase aerodynamic efficiency by approximately 50%.

This paper proposes a remaining useful life (RUL) prediction method that uses the peak of the vibration acceleration signal as an appropriate feature to indicate the degradation process in the rolling element bearings (REBs). In the first step, this feature is transformed into a stationary time series using logarithmic transformation. That is because the long short-term memory neural network (LSTM-NN) works better with the stationary time series. Training the LSTM-NN is performed by this stationary time series as the input and the response is the training time series with values shifted by one time step. Therefore, the LSTM-NN learns to predict the value of the next time step at each point. In other words, to forecast the values of multiple time steps in the future, previous forecasted steps are used as inputs. Next, the values of the future time steps are returned to the main non-stationary form to predict the trend of the peak in the future. Importantly, new measured data can be used to perform new predictions. For this purpose, for every new measured data, the LSTM-NN repeats the mentioned steps and generates a new trend. This algorithm is a trend-dependent method. Therefore, an REB that has a slow degradation stage in its life, which is corresponding to the growth and expansion of defects in REBs, is appropriate to be studied by this algorithm. This method is implemented on two REBs from PRONOSTIA accelerated-life test which have been used by many researchers in the literature. According to the prediction results, the remaining time that peak amplitude trend touches a given threshold is provided. If this threshold is a criterion for the end of life (EoL), this method can be used to determine the RUL. The performance of the proposed method has been evaluated and the presented results are in a good agreement with the experimental data.

The current study aims to identify the equivalent linear dynamics of an autonomous underwater vehicle in the horizontal plane to be able to design an appropriate linear controller. Autonomous underwater vehicles are increasingly being used to provide researchers with a simple, low-cost, and rapid response capability to collect pertinent environmental data. They are fairly stable platforms with little roll and pitch. Nevertheless, dynamic coupling and non-linearities make it a challenging task to perform identi cation process, stability analysis, and control design. Here, for the rst time, the frequency response analysis and the CIFER software, which utilizes strong mathematical algorithms, are employed to solve this problem. Advanced features such as the Chirp-z transform and composite window optimization are also used to extract high-quality frequency responses and best t equivalent transfer functions. After formulating the problem, a frequency sweep input is designed and applied to the rudder controller in the nonlinear simulation and transfer functions for the heading angle and the rate of turn are derived. In addition, these transfer functions are obtained by perturbed equations of motion to be compared with the transfer functions from CIFER. To evaluate the accuracy of the identi ed models, time domain responses from a zig-zag test are compared with the responses predicated by the identi ed model and the linearized model. The results show that this model is in good agreement with the analytical linear model and performs signi cantly better in presence of noise, thanks to the precise spectral functions and windowing technique. The robustness of the proposed method and transfer functions are also assessed by evaluation of the coherence function and altering the window size, frequency bandwidth, and input commands. The results also show that in a speci c frequency range, nonlinear terms are negligible, and the turn rate could be easily predicted by the time derivative of the heading angle.