Capacitor placement is one of the most important methods in loss reduction and improving the voltage profile of distribution systems. Index, GA and Particle Swarm Optimization methods are used for finding the optimal values of the fixed capacitors in the distribution networks with balanced loading for radial configuration. This methodology is tested on a region of the distribution network of the Andhra Pradesh Eastern Power Distribution Corporation and results are presented. Results are very promising. Particle Swarm Optimization gives better reduction in active power loss with lesser value of capacitor sizes than G.A. and Conventional methods.
Optimal selection of size of the conductors in distribution system would result in reduction in capital investment, reduction in energy losses. In this paper, selection of conductors is done based on cost of energy loss and interest and depreciation on capital investment. The main objective is to minimize the real and reactive power losses and also to maximize savings in total cost of conductors, obeying the current and voltage limits. The effectiveness of the algorithm is tested on a part of distribution network of Eastern Power Distribution Corporation of Andhra Pradesh which consists 22 buses, five different types of conductors have been chosen for the analysis.
A novel, efficient multilevel inverter topology has been proposed which generates the seven level output voltage using seven power switches only along with simple and modern pulse width modulation technique. This inverter has offered the advantages of, less component count, reduced switching loss, high efficiency and low cost. Due to the less number of switch components the complexity in generation of pulses is reduced. In this paper the proposed single phase seven level inverter with seven switches topology for grid connected application is compared with the conventional multilevel inverters and existing seven level inverter with ten switches topology. This paper also presents the simple PWM scheme in which, the three reference signals whose offset magnitude is equal to the carrier signal amplitude, is compared with carrier signal to generate the required control signals. This work is carried out in the MATLAB/Simulink environment and simulation results are presented.
This paper describes a neural network based adaptive pattern recognition approach by making a thorough analysis on a power system for estimation of the critical clearing time. A nine bus system is considered for the purpose of transient stability analysis. Faults at five locations are assumed at different instants. Critical clearing times for all five faults at six different loading levels are obtained. Out of thirty cases, 24 cases corresponding to four faults have been used for training the neural network and remaining six CCTs corresponding to the fifth fault at six loading levels obtained by ANN as well as modified Eular method. The same is repeated for all five faults. Nueral network designed with 12 input neurons, 8 hidden neurons and one output neuron. Back propagation technique is used to adjust the weights. Analytical calculations are compared with the values obtained by neural network. Results show that ANN gives accurate results.
Photo-Voltaic systems consisting of long strings are prone to shading effects which result in a decrease in the total power output of the system. The promising solution to this problem is Distributed Maximum Power Point Tracking (D-MPPT) technique where each Photo Voltaic unit is attached to its respective DC-DC converter. Generally, there are two types of D-MPPT architectures: Parallel for high current and series for high voltage applications. There have been some exceptions in the system when the units are joined in series under situations of partial shade. Cross Coupling Effects (CCE) is a term used to describe such exceptions. The coupling effects have been seen with the use of a boost converter. When it is replaced by a SEPIC converter CCE is eliminated, and optimum use of the system can be seen.