International Journal of Industrial Electronics, Control and Optimization
Volume:6 Issue: 3, Summer 2023

In power systems planning, economic load dispatch considering the uncertainty of renewable energy sources is one of the most important challenges that researchers have been concerned about. Complex operational constraints, non-convex cost functions of power generation, and some uncertainties make it difficult to solve this problem through conventional optimization techniques. In this article, an improved dynamic differential annealed optimization (IDDAO) meta-heuristic algorithm, which is an improved version of the dynamic differential annealed optimization (DDAO) algorithm has been introduced. This algorithm has been used to solve the economic emission load dispatch (EELD) problem in power systems that include wind farms, and the performance of the proposed technique was evaluated in the IEEE 40-unit and 6-unit standard test systems. The results obtained from numerical simulations demonstrate the profound accuracy and convergence speed of the proposed IDDAO algorithm compared to conventional optimization algorithms including, PSO, GSA, and DDAO, while independent runs indicate the robustness and stability of the proposed algorithm.

Scene text detection frameworks heavily rely on optimization methods for their successful operation. Choosing an appropriate optimizer is essential to performing recent scene text detection models. However, recent deep learning methods often employ various optimization algorithms and loss functions without explicitly explaining their selections. This paper presents a segmentation-based text detection pipeline capable of handling arbitrary-shaped text instances in wild images. We explore the effectiveness of well-known deep-learning optimizers to enhance the pipeline's capabilities. Additionally, we introduce a novel Segmentation-based Attention Module (SAM) that enables the model to capture long-range dependencies of multi-scale feature maps and focus more accurately on regions likely to contain text instances.The performance of the proposed architecture is extensively evaluated through ablation experiments, exploring the impact of different optimization algorithms and the introduced SAM block. Furthermore, we compare the final model against state-of-the-art scene text detection techniques on three publicly available benchmark datasets, namely ICDAR15, MSRA-TD500, and Total-Text. Our experimental results demonstrate that the focal loss combined with the Stochastic Gradient Descent (SGD) + Momentum optimizer with poly learning-rate policy achieves a more robust and generalized detection performance than other optimization strategies. Moreover, our utilized architecture, empowered by the proposed SAM block, significantly enhances the overall detection performance, achieving competitive H-mean detection scores while maintaining superior efficiency in terms of Frames Per Second (FPS) compared to recent techniques. Our findings shed light on the importance of selecting appropriate optimization strategies and demonstrate the effectiveness of our proposed Segmentation-based Attention Module in scene text detection tasks.

Due to the growth of renewable energies and the need for sustainable electrical energy, AC microgrids (MGs) have been the subject of intense research. Medium voltage MGs will soon have a special place in the power industry. This paper uses a new and effective control scheme for islanded inverter-based medium voltage MGs using the master-slave (MS) technique. The controllers only need local measurements. The designed controls are based on adaptive input-output feedback linearization control (AIOFLC). These controls have a high-performance response; and are robust against some uncertainties and disturbances. The use of the designed control scheme makes the output voltage of distributed generation (DG) sources have negligible harmonics. Besides, the generated voltage and active/reactive powers track their references effectively. The model of the inverter-based DGs is considered in a stationary reference frame, and there is no need for any coordinate frame transformation. The control method presented in this paper can be used for MGs with any number of inverter-based DGs and parallel inverters. The effectiveness of the proposed control scheme is evaluated by simulation in SIMULINK/MATLAB environment and compared with that of feedback linearization control (FLC) and conventional sliding mode control (CSMC).

Multifunctional features of grid-connected inverters can be used for harmonic compensation of local load voltage and grid-injected current. But, in high-power grid-connected inverters, there is a challenge due to low switching frequency. On the other hand, simultaneous compensation of local load voltage and grid-injected current harmonics is an important issue in grid-connected inverters. Using a Unified Power Quality Conditioner (UPQC) at the Point of Common Coupling (PPC), an improved active harmonic compensation method is proposed which is appropriate for high-power low-frequency grid-connected inverters. The UPQC operates as a combination of a negative shunt virtual admittance and a negative series virtual impedance at the PCC. It suppresses the disturbances caused by local load variation and grid impedance change. Using a low-power, high-frequency UPQC, local load voltage and grid-injected current harmonics up to higher-order components are simultaneously compensated despite grid impedance changes and nonlinear local load variations. The control system is designed according to the impedance-based stability criterion to ensure the system's stability. The theoretical results are validated using different case study simulations in MATLAB/Simulink software.

Grinding in a ball mill is a process with high energy consumption; therefore, a slight improvement in its performance can lead to great economic benefit in the industry. The softness of the product of the grinding circuits prevents loss of energy in the subsequent processes. In addition, controlling the performance of a ball mill is a challenging issue due to its complex dynamic characteristics. The main purpose of this article is to use the ground particle size diagram and acoustic signal in ball mill control, and model their relationship based on least squares method. As a result, by extracting useful data from the the acoustic signal, the optimal condition of the ball mill_ in terms of ground particle size and ball mill load (normal, low, high)_ can be achieved. In doing so, this goal, in this article, innovative ideas such as adaptive quantum basis, sparse representation, SVD and PCA-based methods were used. The proposed method has been practically implemented on the ball mill of Lakan lead-zinc processing plant. Also, a prototype of the device was built. The test results show that the optimal load for the studied ball mill is 10t/h. In this case, the ground particle size is 110-120 microns which is ideal for the purposes of this plant. Also, the power spectrum is in the middle frequency band (frequency range of 300-700 Hz). According to the analysis and results, the proposed method will increase the efficiency of the studied ball mill.

Recent scene text detection methods perform superior on benchmark datasets using deep-learning frameworks. In this paper, we re-implement the state-of-the-art text detection method, character region awareness for text detection (CRAFT), which can detect individual characters of scene text images. CRAFT is a character-based detection method with many advantages in detecting complex text by detecting character units and estimating the area between characters, capable of detecting texts of any shape. In the other words, we improve the detection performance of the baseline method, CRAFT, by some modifications in its architecture and proposing a training scheme that takes benefit of the advanced optimizer. The performance improvements of CRAFT are validated on three benchmark datasets: ICDAR2013, ICDAR2015, and COCO-Text. By applying the pre-trained models on COCO-Text, CRAFT shows that it cannot generalize without fine-tuning. We also improve the ICDAR2015 model and evaluate it on benchmark datasets. The evaluation results show improved precision performance compared to the original pre-trained model with fewer iterations and higher accuracy.

In this paper, a novel method for a multi-objective and risk-based optimal reactive power dispatch is proposed. The method includes two main objective functions: technical and economic. The technical objective involves minimizing the risks of voltage instability, voltage deviation, and flow violation, and the economic objective involves minimizing the costs of reactive power generation, active power losses, load shedding, and active power rescheduling. Using these functions and assigning different weighting factors for each sub-objective, the risk of the events or uncertainties to customers or the grid can be managed. In addition, moment matching is used to discretize and create scenarios from continues probability distribution functions of wind speed and electrical energy uncertainties. As the number of uncertain variables increases, so does the number of scenarios and the simulation time. Therefore, the fast-forward selection algorithm is applied to reduce the number of scenarios. To reduce the computational complexity and the number of topological scenarios, a new contingency filtering method based on high-risky events is proposed. A modified multi-objective PSO algorithm based on a hybrid PSO with sine-cosine acceleration coefficients is proposed to find the Pareto front of solutions. The method is implemented on the modified IEEE 30-bus test system. To demonstrate the effectiveness of the proposed method, the results are compared with previously published literature. The results show that risk-based scheduling increases system reliability and cost-effectiveness compared to traditional scheduling.