Recent Advances in Electrical & Electronic Engineering - Volume 9, Issue 2, 2016

The multimedia data under Spark cloud computing platform are composed of many different types of entities, and all kinds of entity attributes are not exactly same. The traditional multimedia data clustering process based on spectral clustering algorithm assumes that multimedia data are composed of independent entities of the same type, and entities are unrelated with the obtained clustering results giving significant error. A multimedia data clustering method based on semi supervised K-means is proposed, and the K-means clustering algorithm is introduced, on the basis of it, the optimal initial clustering center is determined by the iterative method of graph theory. According to ideology similar to tree clustering algorithm, clustering center is expanded based on the maximum distance tree clustering algorithm the similarity is computed between two data objects and between object and cluster, and the clustering of multimedia data is realized under Spark cloud platform. The simulation results show that the proposed method has better speedup ratio, expansion rate and clustering accuracy, and can be applied to the multimedia data clustering under Spark cloud computing platform.

In the process of multimedia network image transmission for image compression processing #140;the current image processing algorithms only rely on the existing node energy for the compression and transmission#140;without considering the digestive problems that image data will transmit itself, leading to image resources selection restriction and poor image compression effect. The multimedia network image transfer compression method based on wavelet distributed algorithm is raised. Each spectrum of image is transformed into two-dimensional discrete 5/3 wavelet transform, four wavelet subs of wavelet transform of different levels in each spectrum are defined as four tensor, non-negative factorization is processed for all wavelets with tensor, which eliminates residual redundancy of image spectral and image space completely. On the basis of this, an energy optimal selection principle is introduced to select the best multimedia network transmission nodes. The distributed processing algorithm is fused to optimize compression process. Simulation experiments show that the compression method of multimedia network image transmission based on wavelet distributed processing algorithms has better performance.

As a data mining tool, time series facilitates better understanding nature of the development process of things and permits forecasting the future values of the process parameters based on the data recorded in a chronological order. ARIMA is one of the general time series models and capable of representing time series which, although not necessary stationary, is homogeneous and in statistical equilibrium. This paper presents the characterization, main methods and problems of the time series by the detailed specific algorithm of software Eviews on analyzing the ARIMA model modeling methods, as well as its specific steps on drafting particular flow chart. Finally, it deals with the Producer Price Index (PPI) collected from the year 1978 to 2013 in China. The statistics related to first 33 years are used to train the models and the 3 past years are used to forecast. This paper constructs two models as ARIMA (1, 1, 1) and AR (1) by using the autocorrelation and partial autocorrelation function of time series, and by comparing with the Akaike information criterion (AIC) and the results of the model test, the ARIMA (1, 1, 1) is chosen as the best model for forecasting. The future value of PPI in the 3 past years shows that ARIMA (1, 1, 1) model has a minor error. It is concluded that a properly performed analysis of time series can be a useful tool for analysis and short-term prediction.

During the research process of multimedia image coding method, due to the large amount of data in multimedia image, the nonlinearity interference is produced in the transmission, storage and processing. When the traditional method is adopted for research, the bandwidth occupied by channel is affected by interference, resulting in no guarantee for the quality of the reconstructed image. In this paper, a multimedia image coding method based on improved fractal theory is proposed. In order to acquire the fractal structure of multimedia image, and resolve affine transformation expression function of multimedia image and its unique fixed point, an iterative function is designed. Through image segmentation, search and coding, the coding process of multimedia image is realized, and the best matching parent block mapping is utilized to replace sub block until iteration convergence, and the decoded multimedia image is obtained. The experimental results show that the improved algorithm for multimedia image coding, can effectively improve the image compression ratio and image coding quality, under the premise of ensuring the image signal-to-noise ratio, and effectively reduce communication amount and calculation of energy consumption of multimedia image.

In the process of network multimedia data compression, due to limited network resources and redundant data transmission, the traditional compression algorithm for network multimedia data based on wavelet transform will waste communication bandwidth and energy, as it greatly reduces the efficiency of data compression, because all network multimedia data need to carry out wavelet transform one by one. Therefore, a compression algorithm for network multimedia data based on improved wavelet transform is proposed. The shortcomings of the traditional compression algorithm for network multimedia data based on wavelet are also analyzed. The traditional wavelet transform process is divided into three stages: decomposition, prediction, and update, which are analyzed in detail. According to the tree-shaped clustering structure of multimedia network, the calculated amount of wavelet transform is dispersed to all the nodes, which calculates wavelet decomposition for all the nodes at the same time. The improved low-frequency coefficients are sent to the upper node to achieve new origin sequence and the even and odd number nodes are decomposed to get secondary wavelet decomposition coefficients, and realize the network multimedia data compression. The simulation results show that the improved algorithm for compressed network multimedia data efficiently reduces the node energy consumption, and shrinks the network delay.

The traditional multimedia image mining based on association rules, has a big flaw in mining time, thus a multimedia image mining method based on fuzzy pixel difference iterative clustering is proposed. In this method, multimedia image index characteristic value is converted into relative membership degree for fuzzy concept index. On the basis of the multimedia image samples and fuzzy pixel difference comprehensive tradeoff metric equations of different types, the Lagrange function is constructed to acquire fuzzy pixel difference iterative clustering loop iteration model. The model is adopted to analyze average divergence inside the cluster and average separation degree between clusters and choose the optimal clustering number value, so as to achieve mining of multimedia image. The results of simulation experiment show that the mining time and energy consumption of the proposed method are better than the traditional method, and the proposed method has a high value of application.

In the scheduling process of massive data in multimedia network, randomness is high, and calculation process is complex, resulting in large amount of calculation during traditional massive data scheduling process. A massive data scheduling model in multimedia network is proposed based on POMDP, to analyze the drawbacks of the traditional model and introduce the general POMDP model. The parameters related to multimedia network are given, and data request of node is as a random event. Through the event-based method, the POMDP mathematical model for mass data scheduling problem of multimedia network is shaped, and the details of state space observation space, action space, state transfer probability and the observation probability and performance function of the POMDP model are described. Once data blocking request reaches to any nodes in the multimedia network, a decision would be made. The simulation results show that, the proposed model has a very high scheduling performance.

This paper presents the design of a micro-strip antenna which adopts coplanar waveguide to feed electrical wave, the antenna and metal radiator floor which are fixed on the same antenna plane to reduce the antenna's size. The designed antenna satisfies the needs of conformal and integrated disposal of the antenna. Antenna structure adopts square heart-shaped metal radiator, which belongs to the gradient line patch micro-strip antenna. Through the gradient width adjusting of metal radiator of heart structure, the antenna reconfigureable radiation characteristics of surface current can be realized, thus the resonance of antenna with electromagnetic wave of different frequency can also be achieved, thus the bandwidth of designed antenna is improved and it meets with the bandwidth demands of RFID (Radio Frequency Identification) and UWB(Ultra wide-band) antenna system.

With the development of digital watermarking technology, digital watermarking protocol is now drawing the attention for protecting copyrights of digital products. However, the copyright protection is fully protected by watermark technology if it is employed a suitable protocol between the rights owner and the customer. Therefore, currently, there are a number of buyer-seller watermarking protocols proposed, however, an anonymous problem and collusion problem are still unsolved completely. Thus, this paper proposes a high security watermark protocol based on iris biometric for resolving these problems. In addition, this paper also produces watermarking generation mechanism which aims to improve the efficiency of the whole mechanism. Finally, the investigation indicates that the proposed buyer-seller watermark protocol takes care of the security concerns of all parties involved, and the proposed scheme could also find an illegal copy of the content, the violators can be traced back.

This paper proposes a Modified Fractional Order PI controller (MFO-PI) for grid connected variable speed Wind Energy Conversion System (WECS). The proposed MFO-PI employs an integer integral action and fractional order integral action, and a supervisor to select the more suitable integral action looking for the working conditions. The parameters of the MFO-PI controller have been tuned using frequency method to realize the isodamping feature. The MFO-PI controller has been used to guarantee maximum power extraction and obtain power factor unity. To investigate the efficiency of the proposed controller, an experimental bench around real time dSPACE1104 card has been carried out. The experimental results demonstrate the superiority of the proposed controller over integer order controllers for a wide range of wind speed.

The devices in the intelligent substation are required to transmit and process information in Smart Grids. On account of the modular design technology, serial port communication becomes the main technique used to connect intelligent modules as its high reliability, flexibility, and ease-of-use features, which is also used in the communication of the data collection board and the conversion board of the three-phase circuit breaker. Cyclic Digital Transmission (CDT) protocol, which is issued by the Ministry of Energy of China, is widely used in the communication of the power system in China. The framework of CDT protocol is composed of the function, the information word, the transfer rules and the frame structure. The CDT protocol that adopts STM32 as the CPU is developed in this paper, so as to achieve highly automated grid scheduling and reliable and stable communication. Some functions and technologies of STM32 such as Timer, USART, DMA, etc., are described in detail. The test results indicate that the serial port reliable communication based on CDT protocol can be realized, which also shows that the data collection board can communicate with the dispatch center via the protocol conversion board.

Background: In traditional temperature protections for dry-type transformers, a transformer is switched off when its temperature is over a high threshold value. Because the actuation time is too long, the transformer will suffer from very high temperatures. In order to overcome the problem, a faster temperature protection based on temperature variation acceleration algorithm is developed. Methods: By studying the temperature change rules and the variation velocities of temperature, the temperature variation acceleration was proposed. The temperature variation acceleration could be calculated by use of three contiguous sampled values. The temperature variation accelerations under working orders are always negative, the temperature variation accelerations under short circuit faults are always positive. Results: The temperature protection based on the temperature variation acceleration correctly and quickly discriminates the short circuit fault with the influence of the thermal time constant and noises of the sensor. Its actuation time is very short and the corresponding temperature of the transformer is low. So it performs much better than the traditional temperature protection not only in accuracy but also in recognition speed. Conclusion: The temperature protection based on the temperature variation acceleration for dry-type transformers acts much faster and more accurately than traditional temperature protections.