Los nuevos mecanismos del desarrollo rural

Department of Electrical and Electronic Engineering, Faculty of Engineering, University of Peradeniya

When an encoded video bit stream is transmitted through an error prone environment, it is subjected to loss of packets. When a packet loss occurs, it will usually drop the whole frame. As a result, the video appears distorted and produce unacceptable video quality, making error control techniques highly desirable.

This paper presents an efficient error correction algorithm to further improve the reconstructed video quality during transmission over error-prone channels. There are many ways to accomplish forward error concealment. Essentially, they all add a controlled amount of redundancy at either the source encoder or the transport encoder. In the proposed method, H.264/AVC codec version Joint Model (JM) 17.0 is integrated with turbo coding which is a powerful forward error correction technique to improve the quality of a video.

Video transmission can be divided into three sectors; Transmission end, Channel and Receiving end. At the transmitter, a JM encoder compresses the input video signal and generates an information bit stream. At the same time, a turbo encoder generates a one parity bit sequence using information bit stream and another parity bit sequence using interleaved information sequence and transmits both parity streams with the information bit stream. Due to impairments of the transmission media, received information bit stream is distorted and different from the data that was transmitted. In the receiver, JM decoder detects this received information bit stream and turbo decoder detects two parity bit sequences. At entropy decoding, if the H.264 data stream is correctly decoded and identified as error free, it will be directly sent for inverse quantization. However, in the event of an erroneous data reception, the error-prone frame will be concealed to some extent using the inbuilt JM decoder error concealment algorithm, which is frame copy, where each pixel value of missing macro block is replaced by the corresponding pixel of the previous decoded reference frame. This is then sent to the turbo decoder for further correction using two parity bit sequences.

The proposed algorithm is tested using a simulated Gaussian channel environment. The experimental results show that the proposed method outperforms the existing method, which is frame copy, in terms of objective quality metrics. At the same time, it becomes more efficient for high-motion sequences because the frame copy algorithm fails to recover lost information from the previous frame due to their high motion. Thus, the proposed algorithm is capable of significantly improving the video quality that has been corrupted by transmission errors.

Proceedings of the Peradeniya University Research Sessions, Sri Lanka, Vol. 16, 24th _{November 2011}

27

Comparison of MPPT Techniques with Regard to Speed of Convergence

S.S.W. Walker, N.D.B. Liyanage, P.A.G.S. Abeynayake, S.G. Abeyratne and K.R.M.N. Ratnayake

Department of Electrical and Electronic Engineering, Faculty of Engineering, University of Peradeniya

In order to extract maximum power available from a solar cell in a photovoltaic system, maximum power point tracking (MPPT) should be activated. Several MPPT techniques have been developed over the past years. In this paper, five Maximum Power Point (MPP) tracking techniques are compared with regard to their convergence speed and complexity of implementation. They are Perturb and Observe (P&O) method, Incremental Conductance method, Three Point method, Constant Voltage Method and Fractional Short Circuit Current Method. The results of the simulation study are shown for a clear sky situation assuming that there is no sudden irradiance or temperature variances. Finally, it has been proved from simulation studies that Constant Voltage method is simpler to implement and faster in MPP tracking compared to the other described methods.

Proceedings of the Peradeniya University Research Sessions, Sri Lanka, Vol. 16, 24th _{November 2011}

28

Cross-Sectional Imaging of Dielectric Materials using Electrical Capacitance Tomography

K.E. Wijethilake, D.B.W. Abeywardana, S.E. Wijethilake, M.G.C.P. Mediwaththe, J.V. Wijayakulasooriya and D. Uduwawala

Department of Electrical and Electronic Engineering, Faculty of Engineering, University of Peradeniya

Cross-sectional imaging of objects is an important area in many applications, especially in medical and engineering fields. Computed Tomography is one such technique that refers to the cross-sectional imaging of an object by using either transmission or reflection of data over the object. This can be achieved using various fields, such as x-rays, ultrasonic waves, magnetic fields, electric fields, microwaves/ RF etc. Since x-rays have properties of straight line propagation and better penetrating ability, x-rays can be considered as the best among them. However, it cannot be used for simple and low cost applications, because of its harmfulness and high cost of experimental equipment. Therefore, a requirement for finding alternative methods for x-rays emerged.

According to the analysis carried out, it was found that Electrical Capacitance Tomography (ECT) can be applied for cross-sectional imaging of dielectric materials. It was developed in the late 1980’s and it is a technique that can be used to obtain the spatial distribution of a mixture of dielectric materials inside a vessel.

In ECT, the object that is needed to be imaged is placed inside the vessel and the relative capacitances between electrodes are measured. Ultimately these capacitance measurements are converted into an image showing permittivity distribution as a pixel- based plot. Since the measured capacitance values depend on the permittivity values of the dielectric material inside the vessel, substances having different dielectric properties can be identified by using this technique.

The system used in ECT may vary depending on the number of electrodes and the sizes of the vessel and the electrodes. The sensor system discussed in this paper has eight electrodes placed along the periphery of the vessel. The permittivity distribution inside the vessel is obtained using Linear Back Projection Algorithm (LBPA). Also, it was shown that iterative and non-iterative reconstruction methods can be used to improve the results further. The Least Square method is used as an iterative method and the Tikhonov Transform is used as a non-iterative method.

Proceedings of the Peradeniya University Research Sessions, Sri Lanka, Vol. 16, 24th _{November 2011}

29

Optimal Capacitor Placement to LV Distribution Network: A Simple Design Approach using Excel Sheet

A. Gajananan1, M.A.R.M.Fernando1, A. Atputharajah1 and C.A.B.Karunarathna2

1_{Department of Electrical and Electronic Engineering, Faculty of Engineering,}

University of Peradeniya

2

Energy Management Branch, Ceylon Electricity Board, Kandy

Shunt capacitors are usually installed in LV distribution networks to reduce power losses and system voltage drops. Recently, the Ceylon Electricity Board installed LV shunt capacitors in some of their selected 11/0.4 kV distribution substations in the Kandy region. This paper proposes a simple design approach in determining the optimal size and the location of such capacitors by using Microsoft Excel spreadsheet software. Two distribution substations namely Siyambalagastenna (SIY) and Mavilmada (MAV) were selected in Kandy city area. The SIY substation was selected for preliminary studies on capacitor placement whereas the MAV substation was used for investigating the proposed design approach.

The daily load patterns were obtained on both substations by using a Fluke 1735 data logger. The active, reactive, and apparent powers, power factor, voltages and current data were recorded for every minute continuously for 24 hours, with and without capacitors, at the transformer end. In the preliminary studies at SIY substation, the measured power values were compared with calculated values and it was found that the error was within 5% during peak hours and 10% during off-peak hours. In the design approach, the distribution network was built-up in Excel spreadsheets by line parameters (line resistance, reactance and distance) and customer loads connected to each pole. The customer loads were obtained first by calculating the average power consumption (from electricity bills during 6 month period) and then by normalising them by measured power at the transformer end. The pole voltages and the power losses were calculated using cumulative power in the distribution network. The line voltage drops were checked by random voltage measurements at the feeder end of the MAV distribution network. The optimal size and capacitor placement were determined while checking reduction of power losses and improvements of voltage drops. It was found that the proposed method provides useful information in designing capacitor size and location.

Proceedings of the Peradeniya University Research Sessions, Sri Lanka, Vol. 16, 24th _{November 2011}

30

Control Algorithm to Optimize Power Balancing in Smart Grid with Voltage Constraints of Distribution Network