Computational Statistics and Data Intelligence

The aircraft integrity rate is a comprehensive indicator reflecting the performance of the aircraft integrity, and it is also a comprehensive evaluation index that reflects the quality of the objectives of the spare parts supply support system. The disadvantage of using the spare parts satisfaction rate or the expected number of stocks as a systematic evaluation indicator is that the indicator only stays in the supply system and does not go beyond the supply system to consider the impact of the supply system on the aircraft or other spare parts. Calculating aircraft uptime assesses the aircraft's ability to continue operation by combining resources over a period of time with the aircraft's combat capabilities. The aim is to determine which spare parts are prioritized for storage in order to obtain optimal aircraft uptime indicators under the condition of funding caps. But for accurate forecasting of spare parts demand, the distribution status of spare parts that generate demand must first be determined. In this paper, combined with the analysis of the statistical probability distribution characteristics of spare parts demand, taking the total cost budget constraint and the maximum spare parts satisfaction rate as the decision-making goal, this paper proposes a model for spare parts storage optimization decision that meets the optimal goal of aircraft integrity, and gives a solution algorithm of marginal analysis. The example application has verified the feasibility and practicality of the model. The model combines resources over a period of time with the aircraft's flight capabilities to assess the aircraft's ability to continue to fly in a steady-state environment. It can plan the uptime of the aircraft according to a given spare parts inventory, or it can calculate the amounts of spare parts required to complete the flight training plan task under a specified performance standard. It has solved the problem of determining which spare parts are prioritized for storage in order to obtain the maximum aircraft uptime under the constraint of limited funds.

The world currently faces the climate change, as a serious environmental issue, which is likely to have major negative subsequence. To encounter this problem, it is supposed that renewable energy systems (RESs) should be one of the top priorities to ensure proper actions are being taken to tackle the global warming problems. As a particular case, solar cells are increasingly being installed around the world either in large and small scales to help environment and eventually reduce the greenhouse gas emission. This all means that design of the solar cells needs to be improved in order to enhance their performance and this improvement cannot be done without considering safety and reliability of the solar cells. In this talk, a solar cell with a photocurrent and a diode is considered to be investigated in terms of reliability analysis issues. For this purpose, the photocurrent equality is used to introduce a new system performance function for the solar cells. This performance function is then employed to set up a reliability analysis problem that aims at investigating the solar cell reliability. A numerical experiment is also provided to represent how this model can be used to enhance the solar cell design. Two of the most stable and efficient existing reliability analysis methods are selected to solve the proposed problems.

Existing permutation tests for the partial coefficient in a multiple linear regression model are mostly stepwise algorithms which are difficult to be analyzed and compared theoretically. In this paper, we propose a unified framework for testing the partial coefficient in a multiple linear regression model. The unified framework provides a theoretical basis for constructing permutation tests for the partial coefficient as well as theoretical justifications why they work. As a result, several permutation tests are proposed based on the unified framework. The unified framework also enables us to compare the performances of the permutation tests theoretically. By showing that the existing permutation tests for the partial coefficient of a multiple linear regression model are special cases of the unified framework, we are able to identify redundant steps and provide simpler algorithms for the existing permutation tests. Recommendations on which methods to use under different situations are provided based on the theoretical and numerical comparisons we have conducted.

This paper briefly introduces a generalized partially function type linear model with interaction terms and applies the model to the study of population mortality, exploring the effects of air quality index, temperature, GDP per capita and number of beds per 1,000 population in hospitals on mortality in selected Chinese cities in 2019. It was concluded that both the air quality index and temperature were positively correlated with the population mortality rate, while GDP per capita and number of beds per 1,000 population in hospitals were negatively correlated with the population mortality rate.

This paper investigates the content related to sleep quality. A generalized partially functional linear model in which the response variable is scalar and the predictor variable is both functional and scalar is constructed by using the sleep quality of 22 volunteers as the response variable, the hourly heart rate as a functional predictor variable, and Wake After Sleep Onset, the Number of Awakenings during the night, and Daily stress Inventory value as scalar predictor variables, where the model exact link function was not specified. Finally, it was concluded that HR was positively correlated with sleep quality from 9 am to about 15 pm and negatively correlated at other times, and that the WASO, Number, and DSI were inversely correlated with sleep quality, which was also largely in line with our expectations.

Paola Bonacini⁶ defined the hyperstar which is a hypergraph with \(m\) 3‐edges such that all of them have two vertices in common and is denoted by \(S^{(3)} (2,m + 2)\). In this paper, we consider the \(S^{(3)} (2,6)\) decomposition and packing of complete 3-uniform hypergraph. Firstly, the necessary condition for the decomposition of the complete 3-uniform hypergraph into an \(S^{(3)} (2,6)\) is \(v = 0,1,2,4,6 \, (\bmod \, 8)\), \(v \ge 6.\) Secondly, according to the recursive constructions, the required designs of small orders are found. For the hypergraphs with large orders, they can be recursively generated by some designs of small orders. Then, it is proved that the above necessary conditions are sufficient except \(v = 6\), that is, there exists an \(S^{(3)} (2,6)\) decomposition of a complete 3-uniform hypergraph if and only if \(v = 0,1,2,4,6 \, (\bmod \, 8)\), \(v > 6.\) Finally, we prove that there exists a maximum \(S^{(3)} (2,6){ - }\) packing, except \(v = 6,7\). And we find the exact value of the packing number \(d(t,T,v)\) for any integer \(v \ge 6\).

Speech is a fascinating phenomenon in the human body. Speech differs from one person to another. Most likely, no two people have the same properties of speech like pitch, formant frequencies, duration, etc. Also, different people speak different languages. All these stated properties make speech/language a non-linear and non-stationary signal, which is challenging to implement for a speech/language recognition system. Here we have tried to design the language recognition of the Dimasa language by first implementing a single-word recognition system. This Dimasa language is a Sino-Tibetan language. It is tonal in nature and spoken by a few local tribes in North Eastern part of India. It is predominant in the Districts of Dima Hasao, Karbi Anglong, Nagaon, and Cachar of Assam. Along with the Dimapur District and Jiribam region of Nagaland and Manipur states respectively. Different techniques of Artificial Neural Networks (ANN) were used to obtain different results. The approach is based on feature extraction techniques Mel-Frequency Cepstrum Coefficients (MFCC). In this work, the popular Levenberg–Marquardt training algorithm, Scaled Conjugate Training Algorithm, and Gradient Descent Propagation Training algorithm were used to train for the pattern recognition neural network used to identify the language and speech.

Scales are a common form in the preparation of questionnaires. Missing data becomes common in scales for a variety of reasons. The traditional methods of dealing with scale data with missing values usually include direct deletion of cases containing missing values, interpolation using single values, multiple interpolation, etc. In recent years, there has been an increasing number of fill methods for various types of missing data, but few scholars have conducted research on fill methods for missing values specifically for scale data. This paper proposes a method to fill in missing values by using the correlation coefficients between the items of the scale, which is called the correlation coefficient method, in view of the fact that the items of the scale are usually correlated to a certain extent. In this paper, we use some of the scale data obtained from the Q City secondary school students' family education project as an example, and use the mean squared error and mean absolute error as evaluation indicators. By comparison with the common median and plurality methods, it was found that filling the scale data using the correlation coefficient method yielded the smallest mean squared error and mean absolute error, with the median method being the second most effective and the plural method the least effective. Therefore, the correlation coefficient method is more effective and suitable for scale data with large correlations.

A sixth order accurate numerical method is proposed for nonlinear integral equations of the Fredholm type. First the problem is discretized and the integral approximated by the classical second order trapezoid rule. Then, the leading truncation error terms of the quadrature rule are approximated. The first derivatives in the error term are approximated using five-point finite difference formula which has fourth order of accuracy, while the third derivative is approximated with a second order finite difference formula. This results to a new leading error term which is of sixth order, hence a sixth order method. Newton algorithm is then designed to approximate the resulting nonlinear system. Numerical examples are used to demonstrate the accuracy and performance of the method. The results are also compared with those from other published works in which the superiority of the proposed method is demonstrated.

Nowadays big data analysis is one of the most demanding frontier technologies. This article develops the information flow research from the regular information space to the wider information measuring space based on the multi-dimension, multi-angle random dynamic information flow research. In stochastic dynamic delivery system, it estimates A kind of general stable multi-dimensional measurable process. The article investigates a lot of precious information in great generalized process. Thus, it makes the optimal estimation of random signals and the optimal recursive filtering equation under this process. Therefore, it presents an efficient mathematic method and a consistent theoretical foundation for more extensive research has been conducted to improve the efficiency of information transmission systems and optimize their management. Because the system analysis of stochastic dynamic signals has strong global search ability and popularization adaptability, the new intelligent optimization algorithm proposed in this paper is becoming one of the most active research directions in intelligence science, information science and artificial intelligence, and has been rapidly popularized and applied in many engineering fields.

If there is a choice function \(f{:}A(D) \to \left[ k \right]\) such that any arc with a common starting vertex and any arc with a common ending vertex do not have the same color in digraph \(D\), then \(f\) is a proper arc coloring. Let \(S^{ - } (u)\) and \(S^{ + } (u)\) denote by the color sets of the incoming arcs and the outgoing arcs of \(u\) separately. If there is a proper arc coloring \(f\) such that the color sets of the outgoing arcs of \(u\) are not as same as the color sets of the incoming arcs of \(v\) for every arc \(\overrightarrow {uv} \in A(D)\), then \(f\) is neighbour-distinguishing. The neighbor-distinguishing arc coloring \(f\) of \(D\) requires the minimum arc chromatic number to be denoted as \(\chi^{\prime}_{ + ,- } (D)\). In this article, we prove the upper bound of a general digraph and the exact results of neighbor-distinguishing arc chromatic number of several special digraph.

Stillbirth is a serious problem globally with approximately 2.65 million third trimester stillborn baby cases recorded around the world each year. In Malaysia, there are very limited studies on the risks of stillbirth. Thus, the objective of this study was to investigate the odds and probability of stillbirth due to maternal demography (race, marital status, maternal age and antenatal care) among women in Malaysia. A case control retrospective study was undertaken at one public hospital in Selangor. A descriptive analysis was conducted to know the percentage of mothers with maternal demography with birth of their babies. Chi-square test of independence was conducted to identify the association between maternal demography and stillbirth. After that, the logistic regression is conducted to investigate the odds and probability of stillbirth due to maternal demography (race, marital status, maternal age and antenatal care) among women in Malaysia. A descriptive result of the variable for stillborn babies indicated that mothers who are ‘Bumiputera’ (66.5%), married (94.2%), age less than or equal to 35 years old (77.4%) and having antenatal care (88.4%). The chi square tests of independence results concluded that only age and antenatal care are associated with birth of baby. Based on the odd ratio it can be conclude that a pregnant woman whose age is greater than 35 years old and with no antenatal care during pregnancy have a higher risk of getting stillbirth baby.

An essential issue concerning the elastohydrodynamic lubrication (EHL) problem in the higher pair contacts of gear and bearing is the viscosity calculation of lubricating oil, which has a non-negligible effect on the EHL characteristic parameters such as film thickness, friction coefficient, temperature rise of oil film and substrate and so on. Roelands empirical formula is commonly used in the calculation of lubricant viscosity, and considered to have higher accuracy than exponential function. However, the operation temperature and pressure of the lubricating oil have exceeded its application range in the contact of the high pair. In this paper, the effectiveness of Roelands formula with constant VPC (viscosity-pressure coefficient) in EHL analysis of higher pairs is investigated. The film thickness and temperature rise of EHL point contact are computed according to the fitted VPC expression varying with the temperature, and the results are compared with those with the constant VPC. The results show that the lubricant viscosity, film thickness and film temperature rise of EHL are highly sensitive to the VPC at low temperature and high pressure. Even a modest change in the VPC produces an important rise/reduction in the above characteristic parameters. The Roelands formula with constant VPC causes a certain deviation to the central film thickness (CFT) and maximum film temperature (MFT) in the EHL point contact when the reference temperature is small, and the deviation decreases gradually with the increasing of the reference temperature.

Low frequency pressure fluctuation is an important topic on the stable and safe working of the system. As the excitation source of the pressure fluctuation, oxygen injection condensation is the indispensable process in the pipeline system of liquid fuel aircraft. Due to the intense mass transfer and flow instability, it is difficult to capture the accurate pressure characteristic of oxygen jet condensation, especially in the first main frequency of pressure oscillation. Aiming at the mechanism and excitation process of pressure oscillation, a numerical simulation is carried out using a modified mass transfer model. The height function method is implemented to calculate the curvature of gas–liquid interface. It can update the mass transfer rate in real-time, which is the core technology to evaluate the pressure fluctuation. The modified model is verified by a water steam jet condensation simulation and the numerical result agrees well with the experimental data. The low-frequency pressure fluctuation characteristic is obtained successfully. The first main frequency is 9.5 Hz with the apparent amplitude of approximately 80 kPa. The research shows that the periodic mass transfer rate and the swing of the continuous oxygen gas plume are the key factor causing the low frequency pressure oscillation. There is an oxygen suck-back flow phenomenon in the oxygen chamber of condenser pipe. In addition, it is also found that decreasing the injection area can restrain the pressure oscillation effectively. These conclusions provide a theoretical guidance for overcoming the low-frequency pressure oscillation eventually.

A novel mechanical bulging process is proposed in this work to overcome the disadvantages of the existing forming process for large rectangular bellows with deep waves. Numerical simulation was conducted on the mechanical bulging process of the three-layer 304 stainless steel rectangular bellow. Simulation results revealed that each layer steel had similar deformation and equivalent strain distribution. The maximum equivalent strain at the round corner of the rectangular bellow was 0.406. The general dimensions of the rectangular bellow were well consistent with the designed values. The wall thickness thinning rate at the round corner was greater than that in the straight edges, which had a maximum value of 20.48%.

Local pulse wave velocity (PWV) is a widely accepted index for quantitatively evaluating the stiffness of regional arteries, its predictive value is important for assessing the risk of various cardiovascular diseases. High frame rates are required to capture the propagation of pulse wave because it travels at high speed. Plane wave technique, in which all transducer elements are excited at the same time for a full aperture transmission of the region of interest thereby creating a high frame rate, and therefore is suitable for local PWV estimation. However, there have been far less systematic studies focused on the effects of the number and transmission angles of plane waves on the performance of coherent plane wave compounding (CPWC)-based local PWV estimation. In this study, the quantitative effects of these two parameters are assessed by a carotid vessel phantom. First, the blood mimicking fluid is pumped through the phantom via CompuFlow1000. And the linear array probe is attached to the Verasonics system for data acquisition of 3 and 5 plane wave sequences with transmission angles ranging from 1° to 8° and an interval increment step of 1°. The local PWV is then estimated by repeatable and robust ultrasonic transit time method The quantitative effects of different configurations on the performance are assessed by mean values and standard deviations of the estimated local PWVs. The results show that 3 plane waves with transmission angles of (−1°, 1°, 0°) is the best configuration by providing the optimal performance for local PWV estimation. The quantitative findings can provide important advice and improve direction for the CPWC-based local PWV estimation.

Vehicle localization is a basic and important part of intelligent and connected vehicle technology. However, intelligent and connected vehicles may be affected by outliers when they are driving in complex environments. Additionally, there may be random packet dropouts and delays when positioning data is transmitted by vehicle-to-vehicle communication technology. These factors can cause intelligent and connected vehicles to be unable to obtain real-time and accurate position information, which poses security risks to intelligent transportation. To ensure accurate positioning of intelligent and connected vehicles, this paper takes into account factors such as packet dropouts, delays, and outliers during data transmission. The measurement packet dropouts and randomly delayed data transmissions are described by two Bernoulli variables, while the outliers are modelled as a student’s t distribution. As a result, a robust state-space model is developed, and the unknown states and parameters of the proposed model are calculated by the variational Bayesian method. Finally, the experimental results indicate the effectiveness.