Search

Abstract This paper proposes a novel approach for improving the speech of a single speaker in noisy reverberant environments. The proposed approach is based on using a beamformer with a large number of virtual microphones with the suggested arrangement on an open sphere. Our method takes into account virtual microphone signal synthesizing using the non-parametric sound field reproduction in the spherical harmonics domain and the popular weighted prediction error method. We obtain entirely accurate beam steering towards a known source location with more directivity. The suggested approach is proven to perform effectively not just in boosting the directivity factor but also in terms of improving speech quality as measured by subjective metrics like the PESQ. In comparison to current research in the area of speech enhancement by beamformer, our experiments reveal more noise and reverberation suppression as well as improved quality in the enhanced speech samples due to the usage of virtual beam rotation in the fixed beamformer. Text for this section.

IoT technology is one of long waves of technology that will act as a platform for countries to rise or fall in the economic path. If countries and companies that take advantage of the opportunities ahead of this technology, they can join the ranks of global leaders and consolidate their position for decades. In recent years, Iran has carried out extensive activities to develop various technologies, but due to the multiplicity of technology fields and the spread of the country's potential in various fields, sufficient accumulation of knowledge and capital in one sector has not taken place. In general, in order for economic coexistence, there must be leaps in several industrial and technological sectors so that a country can become the backbone of a developed country. But since IoT technology is a common technology and is used in all sectors, there is a risk of spreading the scientific and capital potential of the country in its various fields. Therefore, it is necessary to create accumulation of knowledge, capital and wealth by focusing on priority areas. Due to this, in this research, the applied parts of IoT technology that can cause technological and economic leaps have been identified and then the importance and priority of these parts can be determined using nonlinear modeling approach. The results showed that focusing on energy intelligence in Iran has the highest advantage and focusing on this sector can ensure economic growth and technology of society.

As manufacturers and technologies become more complicated, manufacturing errors such as machine failure and human error have also been considered more over the past. Since machines and humans are not error-proof, managing the machines and human errors is a significant challenge in manufacturing systems. There are numerous methods for investigating human errors, fatigue, and reliability that categorized under Human Reliability Analysis (HRA) methods. HRA methods use some qualitative factors named Performance Shaping Factors (PSFs) to estimate Human Error Probability (HEP). Since the PSFs can be considered as the acceleration factors in Accelerated Life Test (ALT). We developed a method for Accelerated Human Fatigue Test (AHFT) to calculate human fatigue, according to fatigue rate and other effective factors. The proposed method reduces the time and cost of human fatigue calculation. AHFT first extracts the important factors affecting human fatigue using Principal Component Analysis (PCA) and then uses the accelerated test to calculate the effect of PSFs on human fatigue. The proposed method has been applied to a real case, and the provided results show that human fatigue can be calculated more effectively using the proposed method.

Nowadays, many accidents, malfunctions, and quality defects are happening in production systems due to Human Errors Probability (HEP). Human Reliability Analysis (HRA) methods have been proposed to measure the HEP based on Performance Shaping Factors (PSFs), but these methods do not have a procedure to select the effective PSFs and consider the PSFs dependency. In this paper, we propose an Artificial Neural Network based Human Reliability Analysis (ANNHRA) in cooperation with Response Surface Method (RSM). This framework uses the advantage Systematic Human Error Reduction and Prediction Approach (SHERPA) method to quantify the PSFs and the ANN and RSM to consider the PSFs dependency and select the most effective PSFs. This framework decreases the time and cost and increases the accuracy of HRA. The proposed framework has been applied to a real case and the provided results show that human reliability can be calculated more effectively using ANNHRA framework.

Examining the trend of the global economy shows that global trade is moving towards high-tech products. Given that these products generate very high added value, countries that can produce and export these products will have high growth in the industrial sector. The importance of investing in advanced technologies for economic and social growth and development is so great that it is mentioned as one of the strong levers to achieve development. It should be noted that the policy of developing advanced technologies requires consideration of various performance aspects, risks and future risks in the investment phase. Risk related to high-tech investment projects has a meaning other than financial concepts only. In recent years, researchers have focused on identifying, analyzing, and prioritizing risk. There are two important components in measuring investment risk in high-tech industries, which include identifying the characteristics and criteria for measuring system risk and how to measure them. This study tries to evaluate and rank the investment risks in advanced industries using fuzzy TOPSIS technique based on verbal variables.

The aim of training is to promote knowledge skills and changing the attitude of the trainers. Therefore, this research is done to evaluate and assess the effectiveness of specialized training courses for locomotive drivers of Iranian Railways in Razavi Khorasan province according to Kirkpatrick's Four-Level Model. This research is applied (in terms of purpose), evaluative (in terms of function) and descriptive (in terms of data gathering method). It is also a survey and linear research, which is calculated in the four levels of reaction, learning, behavior and results. The sample of this research is all the locomotive drivers participated in the courses held in the last 3 months of 2016 and the beginning 3 months of 2017 along their managers and supervisors who were 260 persons. 152 persons are selected as the sample using simple random sampling according to Krejcie-Morgan table. The results obtained from this research show the respondents' reaction to satisfactory educational courses, as their learning average from the training courses raised to 17.23 from 7.94. The change in the behavior of learners after 6 months from the training course was 4.033 and the degree of realization of organizational goals from the perspective of manager was 3.33. According to Kirkpatrick's four-level model, training courses have been considered successful and effective at all levels. It also shows that the organization has achieved its goals by implementing effective training courses.

Acoustic experimental data of IRIB (Islamic Republic of Iran Broadcasting) server room was taken to measure the acoustic noise characteristics of frequency and Sound Pressure Level (SPL). The purposes of this paper are measuring the noise pollution and its evaluation through the regulations. The measurements were taken in two stages. The first one was taken at 10 points by Sound Investigator 2260 using ISO 3741: 1999(E) standard. The results indicate that IRIB server room is an acoustic noise polluted according to Iranian National Building Regulations. Measured data and its frequency analysis show that the high sound energy belongs to low frequencies in the server room. This is confirmed by the next stage of measurement. It was an hour long non-stop noise measurement in the server room to investigate noise exposure level by customers and employees. The Equivalent instant noise level on Aweighting network was 75 dBA. The results are valid for all of data centres, although this study is involved with a special case study. Two main factors are important in server room noise pollution scenario; the amount of sound absorbance in the space and environmental noise characteristics. Most surfaces in server rooms are hard and have low absorbance coefficient, therefore, the intensity of travelling sound increases. On the other hand, the main noise sources in any server room are fans which work usually on a narrowband with a fundamental frequency. So as the main features of server room acoustic noise are the same, the presented results can be generalized to any server room.