Qichang Duan

A Two-Stage Particle Swarm Optimization Algorithm for MPPT of Partially Shaded PV Arrays

ABSTRACT The power-voltage (P-V) characteristic curves of a PV array are nonlinear and have multiple peaks under partially shaded conditions (PSCs). This paper proposes a novel maximum power point tracking (MPPT) method for a PV system with reduced steady-state oscillation based on a two-stage particle swarm optimization (PSO) algorithm. The grouping method of the shuffled frog leaping algorithm (SFLA) is incorporated in the basic PSO algorithm (PSO-SFLA), ensuring fast and accurate searching of the global extremum. An adaptive speed factor is also introduced into the improved PSO to further enhance its convergence speed. Test results show that the proposed method converges in less than half the time taken by the conventional PSO method, and the power is improved by 33% under the worst PSCs, which confirms the superiority of the proposed method over the standard PSO algorithm in terms of tracking speed and steady-state oscillations under different PSCs.

Adaptive fuzzy control used in DFIG VSCF wind power generator system

Wind power generation system is typical multi-variable, non-linear and too complex to build precise mathematical model. In order to capture maximum energy at certain wind speed range and regulate active and reactive power independently, adaptive fuzzy control based on stator flux-oriented vector control was adopted. The simulation program was written in C compile language and Matlab. The results show the good performance of this system and prove the correctness and availability of this scheme.

An intelligent algorithm for maximum power point tracking in photovoltaic system under partial shading conditions

In a photovoltaic (PV) system, maximum power point tracking (MPPT) under partial shading (PS) conditions is a challenging task due to the presence of multiple peaks in the power voltage characteristics. This paper puts forward a novel artificial fish-swarm algorithm (FSA), which is optimized by particle swarm optimization with extended memory (PSOEM-FSA). In this algorithm, both the velocity inertia factor and the memory and learning capacity of PSOEM are introduced into the FSA. To validate the effectiveness of the novel algorithm, the PV system along with the proposed MPPT algorithm was simulated using Matlab/Simulink Simscape tool box. The simulation results show that the proposed approach is effective in MPPT under PS conditions and has a more stable performance when compared with the traditional methods in convergence speed and searching precision.

Comprehensive improvement of artificial fish swarm algorithm for global MPPT in PV system under partial shading conditions

Due to the non-linear characteristics I–V of the photovoltaic (PV) curve, the tracking of the maximum power point (MPP) under partial shading (PS) conditions can sometimes be a challenging task. This paper presents a modified artificial fish swarm algorithm (AFSA) for MPP tracking (MPPT) in PV modules under PS. In this algorithm, the AFSA optimized by particle swarm optimization (PSO) algorithm with extended memory (PSOEM-FSA) is improved by hybridizing it with adaptive visual and step, and the resulting algorithm is a comprehensive improvement on the AFSA (abbreviated as CIAFSA). Combining the searching capabilities of the PSOEM-FSA and the self-learning ability of adaptive visual and step for AFSA, CIAFSA is developed. To validate the effectiveness of this novel MPPT technique, the PV system along with the proposed MPPT algorithm is simulated using the Matlab/Simulink Simscape toolbox. Results show that the proposed approach is more effective in MPPT in PV systems under PS conditions when compared with other methods in searching precision.

Based on Particle Swarm Optimization real-time license plate recognition

The traditional plate character recognition algorithm for the low recognition rate and identify the shortcomings of slow, the paper used PSO algorithm optimization neural network weights and threshold parameters, resulting in greatly improved the license plate character recognition rate and Recognition speed. The experimental results indicate that the PSO algorithm optimized for real-time neural network license plate recognition, the correct identification rate of 99 percent and above, the recognition time is 0.27 s, and the recognition rate and recognition speed is superior to other traditional identification methods, and basically meet the requirements of the application.

Maximum Power Point Tracking for Cascaded PV-Converter Modules Using Two-Stage Particle Swarm Optimization

The paper presents a novel two-stage particle swarm optimization (PSO) for the maximum power point tracking (MPPT) control of a PV system consisting of cascaded PV-converter modules, under partial shading conditions (PSCs). In this scheme, the grouping method of the shuffled frog leaping algorithm (SFLA) is incorporated with the basic PSO algorithm, ensuring fast and accurate searching of the global extremum. An adaptive speed factor is also introduced to improve its convergence speed. A PWM algorithm enabling permuted switching of the PV sources is applied. The method enables this PV system to achieve the maximum power generation for any number of PV and converter modules. Simulation studies of the proposed MPPT scheme are performed on a system having two chained PV buck-converter modules and a dc-ac H-bridge connected at its terminals for supplying an AC load. The results show that this type of PV system allows each module to achieve the maximum power generation according its illumination level without affecting the others, and the proposed new control method gives significantly higher power output compared with the conventional P&O and PSO methods.

Artificial Bee Colony Algorithm Based on Adaptive Search Equation and Extended Memory

ABSTRACT From the perspective of psychology, a modified artificial bee colony algorithm (ABC, for short) based on adaptive search equation and extended memory (ABCEM, for short) for global optimization is proposed in this paper. In the proposed ABCEM algorithm, an extended memory factor is introduced into store employed bees’ and onlooker bees’ historical information comprising recent food sources, personal best food sources, and global best food sources, and the solution search equation for the employed bees is equipped with adaptive ability. Moreover, a parameter is employed to describe the importance of the extended memory. Furthermore, the extended memory is added to two solution search equations for the employed bees and the onlookers to improve the quality of food source. To evaluate the proposed algorithm, experiments are conducted on a set of numerical benchmark functions. The results show that the proposed algorithm can balance the exploration and exploitation, and can improve the accuracy of optima solutions and convergence speed compared with other current improved ABCs for global optimization in most of the tested functions.

Saliency detection based on BP-neural Network

Saliency detection has a significant influence on improving image analysis and processing techniques. We propose a novel saliency detection method based on BP-neural Network in this paper. First, we need to segment images into superpixels by the SLIC approach. Secondly, from these superpixels, we extract 10-dimensional feature to describe a superpixel. Finally, we use BP-neural Network to learn the relationship between 10-dimensional feature value and saliency. We show the results that our proposed method outperform the several state-of-the-art methods on MSRA1000 public database.

Adaptive torque control using RBF neural networks for nonlinear DC chassis dynamometer drive

For the DC chassis dynamometer, a nonlinear mathematical model was established based on the analysis of the transmission system of the DC dynamometer, and an adaptive controller based on RBF NN (radial basis function neural network) was proposed to control a dynamometer to load resistance intelligently to achieve stepless simulation of inertia. By using the Lyapunov synthesis approach, it was proved that the closed-loop system is uniformly ultimately bounded in the presence of bounded neural network approximation error and bounded disturbance force. Simulation results show that the developed controller can offer a good control performance.

Research of fire hazard critical guidelines of mine use belt conveyor and automatic fighting fire system

The fire hazard critical guidelines of coalmine used belt conveyors were researched, and the early parameters of fire hazards were tested. Based on the experiments, relative sensors and a fire detecting system were developed, which made up of sensors and accessories. Again water spray fighting fire technology was researched, and effects of fighting fire were tested and inspected by simulation in experimental roadway.