Abu Zaharin Ahmad

Five-Level Diode-Clamped Inverter With Three-Level Boost Converter

This paper proposes an active front-end solution to balance the dc-link capacitor voltage of the five-level diode-clamped inverter. Capacitor voltage balancing is performed by a three-level boost converter connected to the two inner capacitors of a five-level diode-clamped inverter and additional balancing circuits at the other two outer capacitors. The proposed configuration is tested through simulation and experiment for various load power factor conditions at a high modulation index. The result demonstrates the reliability of the proposed configuration to balance the dc-link capacitor voltage at the desired level.

Advance Two-Area Load Frequency Control Using Particle Swarm Optimization Scaled Fuzzy Logic

One of the most important rules in electric power system operation and control is Load Frequency Controller (LFC). Many problems are subject to LFC such as a generating unit is suddenly disconnected by the protection equipment and suddenly large load is connected or disconnected. The frequency gets deviated from nominal value when the real power balance is harmed due to disturbances.LFC is responsible for load balancing and restoring the natural frequency to its natural position. In this paper, PSO-fuzzy logic technique for Load Frequency Control system was proposed. PSO optimization method is used to tuning the input and output gains for the fuzzy controller. The proposed method has been tested on two symmetrical thermal areas of an interconnected electrical power system. The simulation results are carried out in term of effectiveness of the frequency time response on its damping and compared it to common PID controller. The results show the performances of the proposed controller have quite promising compared to PID controller.

Agents for fuzzy indices of reliability power system with uncertainty using Monte Carlo algorithm

The standard deviation of load level uncertainty in power system reliability assessment has a different value for each load level leading to complexity iterations required in the convergence of Monte Carlo algorithm. In this present work, the fuzzy system agent perspective would be used to control such convergence. Two agents are developed based on fuzzy parameters of Monte Carlo i.e. current with its means and variances; the other agent is the probability of outage capacity for each state. These agents shall be applied in terms of the loss of load probability (LOLP) and loss of load expectation (LOLE) which when implemented and compared based on a Malaysian distribution network (DISCO-Net). The obtained outcomes showed that the fuzzy parameters of Monte Carlo provided a better limitation for variance techniques in uncertainty load levels.

New reliability index for power system protection based on multi-agent technique

A new index reliability for power system protection by optimizing the relay time operating using multi agent system (MAS) is presented. In the proposed method, the MAS consists of two agents as follows; first agent is a fault current agent, i.e., to determine the fault current at all points before and after grounding. While, the second agent is the time operating, i.e., to determine the time operating relay before and after modifying the fault current. The simulation implementation is carried out in Iraqi national grid system (132 kV) - Najaf city. The results show that the failure rate is decreased to 50% approximation for over current and earth fault relays.

Distributed generation location using multi-agent system for reliability assessment

Recently, distributed generation (DG) exploited is predicted to play an important role in the electric power system in term of low cost, security and reliability. In this paper, a new technique for optimize the DG location based on multi agent system (MAS) model for power system reliability evaluation was proposed. The MAS consist of three agents as follows; first agent is a grounding index, which is responsible for adding DG randomly by grounding indices. Second is a reliability evaluation agent that uses a recursive algorithm to predict the suitable generator based on the frequency and duration reliability indices in each state. Third agent is the storage and transfer of data between the other two agents. These agents are created from Malaysia distribution network (DISCO-Net) that implemented by using JADE package software. The simulation results show the effectiveness of proposed multi-agent approach can be taken as a measurement criterion for selecting the best location to add DG in order to improve the power system grounding and reliability.

Load frequency control in microgrid using fuzzy logic table control

This paper presents the load frequency control in isolated ac microgrid using fuzzy logic table control. The objective of load frequency control is to hold the frequency constant against any load changes. The main task of microgrid is to maintain the voltage level supply to the loads. Any mismatch between the power generations and loads create to the major problem of frequency fluctuation that effected to the active power sharing and reliability in microgrid power flow. In this paper, the model of the generation unit in microgrid is carried out for proposing the load frequency control using fuzzy logic table. The comparison of the proposed method is compared with the common proportional integral (PI) controller. The simulation results have shown that the proposed method is better than PI controller in term of transient performance. The simulation is carried out through the Matlab/Simulink environment.

Hybrid micro-grid control and active power sharing using MIT rule based on speed droop controller

This work investigates dynamic behavior and improves the performance of the active power sharing. The transfer of micro-grid from grid-connected mode to island mode is faced by many challenges. In addition, the minimizing frequency fluctuation during sudden changes in the load is still an issue. A new proposed method to control the grid frequency for active power sharing is presented using adaptive MIT rule. The comparison of the response is then compared with the conventional PI controller for the output response verification. The results show the proposed method producing promising and better response. Micro-grid model is developed using Matlab/Simulink.

Power energy management strategy of micro-grid system

In this paper, the power energy management of the micro-grid system that consists of photovoltaic (PV), wind and energy storage systems is analyzed. The micro-grid is proposed to cater the load demand in standalone mode. The operation is set the PV to act as a primary source follow by the wind energy and the energy storage as a back-up source. The main focus in this paper is to propose the strategy of energy management of renewable energy sources. The design consideration and analysis are setup in Matlab/Simulink environment.

Comparative study of short prediction MPC-KF with PI controller for DC-DC buck converter

In nature, the DC-DC converter deals with high switching phenomenon that contributes to the higher conduction losses. Its operation conventionally associated with PI/PID controller in order to meet the desired output. To increase the DC-DC converter performance, an advanced controller has been proposed. However, it is deals with complex algorithm that made the controller cost higher. In this paper, the short prediction horizon of MPC using search tree optimization that generates low switching phenomenon is proposed. The optimal problem is being solve using on-line to reduce the computational effort. The Kalman filter is then used for load estimation. This control approach for DC-DC converter has produced the promising output transient performance when compared with conventional PI controller while also minimizing the switching losses phenomenon.

A new parameters identification of single area power system based LFC using Segmentation Particle Swarm Optimization (SePSO) algorithm

The accuracy of the power system model is important in investigating the transient phenomena of load frequency control (LFC). In this paper, Segmentation Particle Swarm Optimization (SePSO) method is proposed for governor-turbine model identification of single area power plant. The method is acquired based on a combination of segmentation and Particle Swarm Optimization (PSO) algorithms, in which the segmentation is used to recognize the local and global optimal point problem of PSO. The tests of the investigated governor system performed to obtain the step response to identify all parameters of the governor-turbine model. Finally, to verify the effectiveness of the proposed identification, three disturbance cases and three control parameters were implemented by comparing with the common PSO and GA-PSO algorithms. The results show that the proposed method performed better in terms of accuracy and computation time.