فصلنامه روشهای هوشمند در صنعت برق
پیاپی 27 (پاییز 1395)

Abstract: Speech is the basic mean of communication among human beings.With the increase of transaction between human and machine, necessity of automatic dialogue and removing human factor has been considered. The aim of this study was to determine a set of affective features the speech signal is based on emotions. In this study system was designs that include three mains sections, features extraction, features selection and classification. After extraction of useful features such as, mel frequency cepstral coefficient (MFCC), linear prediction cepstral coefficients (LPC), perceptive linear prediction coefficients (PLP), ferment frequency, zero crossing rate, cepstral coefficients and pitch frequency, Mean, Jitter, Shimmer, Energy, Minimum, Maximum, Amplitude, Standard Deviation, at a later stage with filter methods such as Pearson Correlation Coefficient, t-test, relief and information gain, we came up with a method to rank and select effective features in emotion recognition. Then Result, are given to the classification system as a subset of input. In this classification stage, multi support vector machine are used to classify seven type of emotion. According to the results, that method of relief, together with multi support vector machine, has the most classification accuracy with emotion recognition rate of 93.94%.

With rapid development of digital technology, protecting information such as copyright, content ownership confirmation has become more important. In image watermarking, information of the image is inserted such that the visual quality of the image is not reduced and the receiver is able to get the required information. Some attacks such as image cropping, destroy the watermark’s information. In this article, a new watermarking scheme is proposed which is robust against tough cropping. In the proposed scheme, classic Sudoku table which is a 9*9 table, has been used. One feature of Sudoku table is that Sudoku's limitations cause uniform scattering of symbols or numbers throughout the table. In the proposed scheme, Sudoku table and both watermarking approaches based on spatial domain and transform domain such as DCT and DWT are used. Lack of using of soduko solution at the stage of extraction and finding correct solution to obtain watermark, is innovation of this scheme. Robustness of watermarking against cropping attack is up to 92%, which shows good and effective performance of the proposed scheme.

The classic spectrum management methods cannot efficiently supply the future multicast traffic demands. Recently, the cognitive radio technology has emerged as a promising solution to enhance the spectrum utilization. However, the potential heterogeneity in channel availability among the network users leads to send the packets through multiple channels. This significantly decreases the network throughput. The main goal of this paper is to reduce the multicast period in a single-cell and multi-channel wireless mesh cognitive radio network. In this regard, we jointly exploit the multicast scheduling policies, the intra-groups and inter-groups assistance strategies, and the Wireless Broadcast Advantage (WBA). Based on WBA, a single transmission of a node can simultaneously cover multiple neighboring receivers. In addition, by assigning multiple radios to each network user, we allow the nodes to simultaneously transmit/receive the packets on distinct channels. Numerical results of our comprehensive simulations demonstrate the efficiency of the proposed scheme in term of the network throughput.

In this paper, a single switch transformerless high step up dc-dc converter with low voltage stress on the switch is proposed. In the proposed converter only one switch is used which makes the control scheme simple as well as reducing the switching power loss. The voltage gain of the proposed converter is higher than the conventional boost converter and buck boost converter and Proposed converter works in wide rang than conventional converters. The proposed converter has low voltage stress on the switch which makes reducing the switching power loss. The proposed converter can be operated in the continuous conduction mode (CCM) and the discontinuous conduction mode (DCM). In this paper, different operation modes of the proposed converter, calculation of the voltage gain, the currents that flow through the components, efficiency and capacitors voltage ripple are presented. To verify the operation of the proposed converter, simulation results via PSCAD software and experimental results are provided.

Implantable multi-channel neural recording Microsystems comprise a large number of neural amplifiers, that can affect the overall power consumption and chip area of the analog part of the system.power, noise, size and dc offset are the main challenge faced by designers. Ideally the output of the opamp should be at zero volts when the inputs are grounded.In reality the input terminals are at slightly different dc potentials.The input offset voltage is defined as the voltage that must be applied between the two input terminals of the opamp to obtain zero volts at the output. Amplifier must have capability to reject this dc offset. First method that uses a capacitor feedback network with ac coupling of input devices to reject the offset is very popular in designs.very small low-cutoff frequency.The second method employs a closed-loop resistive feedback and electrode capacitance to form a highpass filter.Moreover,The third method adopts the symmetric floating resistor the feedback path of low noise amplifier to achieve low-frequency cutoff and rejects DC offset voltage. .In some application we can use folded cascade topology.The telescopic topology is a good candidate in terms of providing large gain and phase margin while dissipating small power. the cortical VLSI neuron model reducing power consumption of circuits.Power distribution is the best way to reduce power, noise and silicon area. The total power consumption of the amplifier array is reduced by applying the partial OTA sharing technique. The silicon area is reduced as a benefit of sharing the bulky capacitor

with release of a huge amount of news on the Internet and the trend of users to Web-based news services.it is necessary to have a recommendation system. To grab attentions to news, news services use a number of criteria that called news values and user location is an important factor for it. In this paper, LONEF is proposed as a tow stage recommendation system. In first stage news are ranked by user’s locations and in second stage news are recommended by location Preferences, recency, Trustworthiness, groups priorities and popularity. To reduce ambiguity these properties is used tow Mamdani fuzzy interference and case-based decision systems. In Mamdani fuzzy interference system, it is tried to increase the system speed by optimizing selection of rules and membership functions and because of ambiguous feedback implementation, a decision making system is used to enable better simulation of user’s activities. Performance of our proposed approach is demonstrated in the experiments on different news groups.