In this paper, a new predictive direct torque control (DTC) method is proposed, improving the dynamic response of the classical DTC and reducing the torque- and flux-ripples through a voltage vector with an optimal phase. In the transient state, the voltage vector phase is selected in a manner where the fastest dynamic response is achieved, while in the steady state, this phase is selected in a manner where the stator flux amplitude reaches its commanding value at the end of the control cycle. In the steady state, the selected vector is applied to the motor with an optimal time duration calculated to achieve the minimum torque ripple. The five-segment space-vector modulation is used to synthesize the voltage vector, where a fixed switching frequency is obtained. To investigate the effectiveness of the proposed method, steady-state and transient-state performances are tested in the MATLAB software and in practice. Both the simulation and experimental results confirm that the proposed method reduces the torque and flux ripples effectively while improving the dynamic response of the classical DTC method. The comparative investigation of the proposed method with the recent DTC methods indicates that the proposed method has lower ripples in the steady state and a faster dynamic response in the transient state.

In this paper, a new predictive direct torque control (DTC) method is proposed, which improves the dynamic response of the conventional DTC in the transient state and yields the minimum torque and flux ripples in the steady state through an optimal voltage vector. At the steady state, the magnitude, phase, and time duration of the voltage vector are adjusted in a manner where the minimum torque and flux ripples are obtained; whereas in the transient state, the voltage vector parameters are adjusted in a manner where the fastest dynamic response is achieved. The space-vector modulation is used in synthesizing the selected voltage vector where a fixed switching frequency is achieved. In order to improve the control system efficiency, the principle of maximum torque per ampere is adopted in obtaining the commanding stator flux magnitude. To investigate the effectiveness of the proposed method, the steady-state and transient-state performances are tested in MATLAB software and in practice. The simulation and experimental results confirm that the proposed method yields the minimum torque and flux ripples while improving the dynamic response of the conventional DTC. The comparative investigation with an existing predictive DTC method indicates that the proposed method has a better performance in both the steady state and transient state.

Building extraction from high-resolution satellite images (HRSI) in urban areas is an intricate problem. Recent studies proposed different methods during 2005–2015. However, in HRSI, they have not investigated the effects of challenges altogether. This paper studies the effects of non-building features which are the main drawbacks in building extraction. To overcome each challenge, it reviews recent strategies between 2005 and 2015. The pros and cons of each strategy are discussed, and proper strategies are combined to generate hybrid methods. Lower cost and fewer strategies are efficient attributes to recognize the best hybrid methods. Hybrid methods can be useful for different case studies in the future.

Despite significant advances and innovations in deep network-based vehicle detection methods, finding a balance between detector accuracy and speed remains a significant challenge. This study aims to present an algorithm that can manage the speed and accuracy of the detector in real-time vehicle detection while increasing detector speed with accuracy comparable to high-speed detectors. To this end, the Fast-Yolo-Rec algorithm is proposed. The proposed method includes a new Yolo-based detection network and LSTM-based position prediction networks. The proposed semantic attention mechanism in the spatial semantic attention module (SSAM) significantly impacts accuracy and speed on par with the most recent fast detectors. Recurrent position prediction networks, on the other hand, improve the detection speed by estimating the current vehicle position using vehicle position history. The vehicle trajectories are classified, and the LSTM network for the specified route predicts the vehicle positions. The Fast-Yolo-Rec algorithm not only determines the position of the vehicle faster than high-speed detectors but also allows for the speed control of the detection network with acceptable accuracy. The evaluation results on a large Highway dataset show that the proposed scheme outperforms the baseline methods.

Histogram which can be modeled as a mixture of two Gaussian density functions, estimating these densities inpractice is not simply feasible. The objective of this paper is to use adaptive particle swarm optimization (APSO) forthe suboptimal estimation of the means and variances of these two Gaussian density functions; then, the computationof the optimal threshold value is straightforward. The comparisons of experimental results in a wide range of complexbimodal images show that this proposed thresholding algorithm presents higher correct detection rate of object andbackground in comparison to the other methods including Otsu’s method and estimating the parameters of Gaussiandensity functions using genetic algorithm (GA). Meanwhile, the proposed thresholding method needs lower executiontime than the PSO-based method, while it shows a little higher correct detection rate of object and background, withlower false acceptance rate and false rejection rate.