سری زمانی

Coronavirus disease 2019 or Covid-19, which is also called acute respiratory disease NCAV-2019 or commonly called corona, is a respiratory disease caused by acute respiratory syndrome coronavirus-2. Forecasting the number of new cases and deaths during todays can be a useful step in predicting the costs and facilities needed in the future. This study aims to model and predict new cases and deaths efficiently in the future. Nine popular forecasting techniques are tested on the data of Covid-19 in Yazd city as a case study. Using the evaluation criteria of mean square error (MSE), root mean square error (RMSE), mean absolute error (MAE), and the mean absolute percentage of error (MAPE) of the models are compared. According to the selected evaluation criteria, the results of the comprehensive analysis emphasize that the most efficient models are the ARIMA model for predicting the cumulative cases of hospitalization of Covid-19 and the Theta model for the cumulative cases of death. Also, the autoregressive neural network model has the worst performance among other models for both hospitalization and death cases.

Understanding and analyzing spatial-temporal data changes is very important in various applications, including the protection and development of natural resources. In the past studies, Markov process and comparison-based methods were mainly used to predict the changes of vegetation indices, whose accuracy still needs improvement. Although time series analysis has been used to predict some indices, the method to extract these indices from remote sensing data and model their sequences with deep learning is rarely observed. In this article, a method for predicting changes in plant indices based on deep learning is presented. The research data includes Landsat satellite images from 2000 to 2018, related to four seasons in the north and east of Shahrood city in Semnan province. The time span of the extracted images makes it possible to predict changes in vegetation cover. Vegetation indices extracted from the data set are NDVI, SAVI and RVI. After performing atmospheric corrections on the images, the desired indicators are extracted and then the data is converted into a time series. Finally, the modeling of the sequence of these data is performed by the Short-Long-Term Memory (LSTM) network. The results of the experiments show that the deep network is able to predict future values with high accuracy. The amount of the model error without additional data is 0.03 for the NDVI index, 0.02 for the SAVI index, and 0.06 for the RVI index.

Subsidence is a very destructive and dangerous phenomenon that, in addition to endangering human life, can also damage the infrastructure of cities. One of the reasons for its creation is the uncontrolled extraction of groundwater, which occurs widely in the plains of Iran. The time Series InSAR method is one of the important methods for accurate and continuous monitoring of subsidence. But the main problem with this method is the removal of pixels with low correlation in the processing cycle. In this study, to overcome this problem, subsidence of Rafsanjan plain has been investigated using the improved SBAS time series method based on coherence. The data used are 15 images of SENTINEL-1 satellite related to the period from October 2015 to October 2016 and 50 interferograms are generated. The results show the ability of this method to use all pixels of the interferogram, even pixels related to vegetation areas with low correlation. The highest subsidence rate was 284 mm per year for Rafsanjan-Bahrman plain and 252 mm per year for Rafsanjan-Kashkoyeh plain along the satellite line of sight. To investigate the relationship between the improved SBAS results and the water level of wells in the region, Pearson correlation coefficient was used, and to model the relationship, a linear regression model was used. The results indicate a strong direct linear relationship. Also, the linear regression model has the ability to model the relationship with a 95% confidence level. Analysis of variance (ANOVA) was used to test the significance of the linear regression model and Durbin–Watson test was used to evaluate the autocorrelation in the residuals. The results confirm the significance of the model and the independence of the observations.

Electric power and power distribution are prominent infrastructures for economic development in any developing country like Iran. Also, the power distribution network is a very important supply chain that combines a variety of processes. Smart electrical energy distribution networks are one of the latest technologies in the world. The main goal of these networks is to provide reliable electricity, increase the reliability factor and network stability, and respond to the growing needs of customers with minimal damage to the environment, profit, and high efficiency. In the last three decades, the rapid evolution and prevalent adoption of information systems, distribution analysis tools, computational models, and more recently, the emergence of smart grid technologies have given utilities access to the data and tools required for improving these analyses and the possibility of increasing the efficiency of power distribution systems (by, for example, reducing losses and optimizing voltage profiles). Forecasting the future state of the network with the least error brings us closer to the smart network. Because electricity is a mortal product, a comprehensive approach to unplanned power extinction (outage) time is very valuable in preventing any power distribution losses. Various accidents disrupt (cause breakdowns in) the power distribution network, which can be repaired and restored without a hotline. One of the main reasons for customers' power outages is the blackouts in the distribution field, which are affected by technical and non-technical events in the electricity distribution networks. Forecasting these events and managing them can be effective in reducing unplanned power extinction (outage) time. The purpose of this article is to present a model for predicting the duration of unplanned power extinction (outage) and unsold energy based on the data recorded from 121 systems (controllers), the urban network of the three power distribution companies in Yazd province. The final result shows that the ARIMAX model(s) shows less error than the ARIMA model(s) and presents better prediction. Therefore, using exogenous variables in predictions and not being satisfied with the fluctuations of a variable can improve predictions. The model proposed for predicting unsold energy is ARIMAX(1,0,1)(0,0,0) considering the number of incidents and the time of unplanned outages as exogenous variables. The model also shows that in July 2022, the unplanned power extinction (outage) time of this network will be approximately ten hours and also the unsold power will be approximately 6 MWH. On the other hand, the community is without electricity and dissatisfaction has arisen, which lies in economic and social losses. Therefore, with this warning, managers should re-examine the factors of disruption and lack of electricity supply and think of measures to reduce these blackouts when planning for this month of the year

Risks in the energy sector are usually very low in frequency but very high in intensity, which has made the energy sector of any country a high-risk sector. Therefore, to cover its risks, appropriate insurance policies must be designed and implemented. The design of such insurance policies requires accurate identification of all risks, determination of their acceptance criteria, and finally doing actuarial calculation for the designed product. This article just focuses on insurance for crude oil and gas refineries in operation. First, based on the literature review, the risks in this field are identified and then, based on the opinions of health and safety experts and risk assessors working in the field of energy, the model for assessing such risks develops in the form of a risk assessment matrix. A Bayesian structural equation modeling using the Amos has been employed to study impact of the presented variables. Then, referring to the information available in the insurance industry, the pricing of this insurance product is presented based on the rate and the sum insured.

In this research, 12 approaches were proposed to create an optimal vector based on supporting vector machine and neural networks based on genetic algorithm, cuckoo and particle swarm Optimization (PSO). In this regard, we have tried to design a system that reduces the cost of data collection. For this purpose, three data sets with time series capability of standard UCI data were used in this study. The results of the approaches used in this research show the good performance of all the used algorithms. However, the ability and performance of each approach vary according to the type and nature of the data. This has sometimes led to better results from the MLP neural network and the GA or Cuckoo algorithm, and in some cases, the PSO algorithm has better outcomes. Regarding the results, it can be said that the use of feature selection based on semi-regulatory classification reduces system error, increases the accuracy and increases the speed of time series estimation. Hence, by using the efficient and powerful MLP Neural Network and backup vector machine along with the optimization algorithm and metamorphic, an optimal combination classification system can be designed for time series estimation.

It is difficult to apply the real world’s conceptions due to their uncertainty. Generally, time series are known to be non-linear or non-stationary. Regarding these two features, a system should be sensitive enough to apply the unity of time series and repeat this sensitiveness in the prediction. A predict system can exactly scrutinize the hidden features of time series and also can have high predicting runs. Lots of statistical tools such as regression analysis, gradient average, exponential gradient average and auto regression gradient average are used in traditional predictions. One of the biggest challenges of these approaches is the necessity of greater observations and the avoidance of linguistic variables or subjective experts’ ideas. Also these methods are limited to linear being assumptions. In order to dominate the limitations of traditional methods, many researchers have utilized soft computations like fuzzy logic, fuzzy neural networks, evolutionary algorithms and etc.
In this paper, we proposed a new fuzzy prediction novel based on the high order fuzzy time series. Our proposed model is based on the higher order fuzzy time series prediction computational approach. In this method a group of features are evaluated, by adding the value of the preceding element of predicting element to the result of the series’ differences. At that, particle swarm optimization is used to optimize Calculation algorithm features, which renders a better performance in order to solve the problems of higher order fuzzy time series. Then by choosing the best features, a result can be inferred as the predicting value.
The performance of the approach is presented in which after the fuzzification of time series and creating the logical fuzzy relations, by using the lower limit of the predicting element’s range and its consecutive range, and the resulted difference of sequential elements, some specific computations are done and a set of features are gained. Then, using the particle swarm optimization function, the best parameter is selected. The fitness function in the proposed method has two parts: a general section (the average of all orders) and a partial (Every columns orders). In general section, the overall average of error is shown. In Every columns orders section each column individually considered. For the second to tenth order (9 PSO separate) the answer is checked. The method is as follow; we used two parameters b and d for the feature calculation algorithm. The amount of d was manually and randomly between 3 – 1000, but PSO find the amount of b.
Properties obtained by this method, have less outliers data and waste, which it causes predicted closer, with less error.
Finally, defuzzification is performed. The yielded score is the predicted integer value of considered element.
In order to decide the precision of the prediction’s rate, we compare the proposed model to other methods using the mean square error and the average error. In order to show the efficiency of the proposed approach, we have implemented this method on the Alabama University’s enrollment database. It can be observed that the suggested method provides better results compared to the other methods and also renders a lower error.

In Air conditioning systems (HVAC), including heating, cooling and ventilation systems use much of the total energy consumed in commercial buildings. Air Handling Unit (AHU) is one of the most important parts in air conditioning systems. In a survey of the UK buildings, data has shown that 25-50% of energy is wasted due to errors in air-conditioned buildings. If these errors are identified in the early stages before unacceptable damage occur and recognized, this range can be reduced to below 15%.
Along with the evolution of energy-efficient air conditioning in commercial buildings, a wide range of diagnostics and fault detection methods continuously increased in the fields of air conditioning. Different techniques are developed using computer with the low installation cost for years to increase real-time detection of errors.
In this study, a method of fault recovery in temperature sensor of air handling unit has been proposed using simulation of field programing analog array (FPAA) and genetic algorithms as evolutionary algorithms. It is assumed that multiple sensors are working and the output is the average temperature sensors. In faulty situation, using the fault recovery method, the system is able to work properly with one of the sensors. The study consists of two parts, which is a part related to bias, scale and drift fault detection using time series method, noise fault detection using the sample standard deviation, constant fault detection using a voter and the another part is fault recovery using FPAA.

Atrial Fibrillation is a supra ventricular tachyarrhythmia, which is characterized by the deterioration of atrial mechanical function and aberrant. It has become a social and economic problem because a large percentage of the world population suffering from this disease. The early diagnosis of this fatal cardiac Arrhythmia can be prevented and managed it. In this study, we used non-invasive methods based on nonlinear analysis of ECG signal, to identify individuals prone to Atrial Fibrillation. This study consists of three steps, the first step is preprocessing of ECG signals collected from the Physionet to removed 50 Hz and Artifacts noises, then the HRV signal is extracted from them. Secondly, extract nonlinear parameters from HRV signal, including the Poincare Plot Deviations, Correlation Dimension, Lyapunov Exponent, Approximate Entropy and Sample Entropy, and the five characteristics of Recurrence Plot Quantitative Analysis. Third we use a Support Vector Machine classifier for the classification healthy people and those at risk for Atrial Fibrillation Attack. Final results show that Support Vector Machine classifier is able to differentiate healthy from patient population with 93% accuracy, when HRV signal analysis 5 minutes before the onset of Atrial Fibrillation episodes.