Minimum Absolute Error Research Articles

How Many Visual Fields Are Required to Precisely Predict Future Test Results in Glaucoma Patients When Using Different Trend Analyses?

To evaluate the minimum number of visual field (VF) tests required to precisely predict future VF results using ordinary least squares linear regression (OLSLR), quadratic regression, exponential regression, logistic regression, and M-estimator robust regression model. Series of 15 VFs (Humphrey Field Analyzer 24-2 SITA standard) were analyzed from 247 eyes of 155 open-angle glaucoma patients. Future point-wise (PW) VF results and mean VF sensitivities were predicted with varying numbers of VFs in each regression method. In PW-OLSLR, as expected, the minimum absolute prediction error was obtained using the maximum number of VFs in the regression (14 VFs); mean absolute prediction error was equal to 2.4 ± 0.9 dB. Ten VFs were required to reach the 95% confidence interval (CI) of the minimum absolute prediction error. Prediction errors associated with the exponential and quadratic regression models were significantly larger than those from PW-OLSLR, whereas errors from logistic regression were not significantly smaller than those from PW-OLSLR; however, the absolute prediction error from the M-estimator robust regression model was significantly smaller than those associated with PW-OLSLR (P < 0.01, paired Wilcoxon test). Like PW-OLSLR, 10 VFs were needed to obtain the minimum absolute prediction error of mean VF sensitivity, but there were no significant differences in errors using the different regression methods. Approximately 10 VFs, are needed to achieve an accurate prediction of PW VF sensitivity and mean sensitivity. Prediction error of PW VF sensitivity can be significantly minimized using the M-estimator robust regression model compared with conventional OLSLR.

A Practical Split-Window Algorithm for Retrieving Land Surface Temperature from Landsat-8 Data and a Case Study of an Urban Area in China

This paper proposes a practical split-window algorithm (SWA) for retrieving land surface temperature (LST) from Landsat-8 Thermal Infrared Sensor (TIRS) data. This SWA has a universal applicability and a set of parameters that can be applied when retrieving LSTs year-round. The atmospheric transmittance and the land surface emissivity (LSE), the essential SWA input parameters, of the Landsat-8 TIRS data are determined in this paper. We also analysed the error sensitivity of these SWA input parameters. The accuracy evaluation of the proposed SWA in this paper was conducted using the software MODTRAN 4.0. The root mean square error (RMSE) of the simulated LST using the mid-latitude summer atmospheric profile is 0.51 K, improving on the result of 0.93 K from Rozenstein (2014). Among the 90 simulated data points, the maximum absolute error is 0.99 °C, and the minimum absolute error is 0.02 °C. Under the Tropical model and 1976 US standard atmospheric conditions, the RMSE of the LST errors are 0.70 K and 0.63 K, respectively. The accuracy results indicate that the SWA provides an LST retrieval method that features not only high accuracy but also a certain universality. Additionally, the SWA was applied to retrieve the LST of an urban area using two Landsat-8 images. The SWA presented in this paper should promote the application of Landsat-8 data in the study of environmental evolution.

Range limited double-thresholds multi-histogram equalization for image contrast enhancement

Thresholding segmentation method is the commonly applied technique for extracting objects from the background of the image. If the single object in the image is clearly distinguishable from the background, the histogram of the image would be bimodal and the threshold can be easily chosen at the bottom of the histogram valley. However, histograms are not always bimodal. Thus, new methods are required to solve this problem. This paper raises a novel contrast enhancement method called range limited double-threshold multi-histogram equalization (RLDTMHE). First, we deduce the Otsu’s double-thresholds method, and divide the image into three parts with the double thresholds. In order to preserve the brightness, range of the equalized image is calculated to yield minimum absolute mean brightness error (AMBE) between the output image and the original one. Finally, each sub-image is equalized independently using the new range. Experiment results show that our algorithm obtains more clear details, while keeping the brightness of the original image very well.

Equivalent transfer function based multi-loop PI control for high dimensional multivariable systems

A new equivalent transfer function (ETF) parameterization algorithm to incorporate the loop interaction effect into the design of multi-loop PI controllers for high dimensional multivariable processes is presented in this paper. The design scheme consists of two stages. In the first, by exploiting the relationship between the equivalent closed-loop transfer function and the inverse of open-loop transfer function, the analytical expression of ETF is derived. In the second stage, based on the ETF, controller parameters for each loop are determined by utilizing the existing PI tuning rules and the simple internal model control method. The proposed ETF parameterization algorithm is more accurate and reasonable compared to the conventional ETF model approximation methods. Furthermore, the advantage of the multi-loop PI controller designed by the proposed ETF is more significant when applied to higher dimensional processes with complicated interaction modes. Several typical industrial process examples show the well-balanced and robust response with the minimum integral absolute error.

基于GOCI影像的湖泊悬浮物浓度分类反演

PDF HTML阅读 XML下载 导出引用 引用提醒 基于GOCI影像的湖泊悬浮物浓度分类反演 DOI: 10.5846/stxb201411152264 作者: 作者单位: 南京师范大学虚拟地理环境教育部重点实验室,南京师范大学虚拟地理环境教育部重点实验室,南京师范大学虚拟地理环境教育部重点实验室,南京师范大学虚拟地理环境教育部重点实验室,南京师范大学虚拟地理环境教育部重点实验室,南京师范大学虚拟地理环境教育部重点实验室,南京师范大学虚拟地理环境教育部重点实验室,南京师范大学虚拟地理环境教育部重点实验室 作者简介: 通讯作者: 中图分类号: 基金项目: 国家自然科学基金项目(41171269,41471282) Method for estimating the concentration of total suspended matter in lakes based on goci images using a classification system Author: Affiliation: Key Laboratory of Virtual Geographic Environment,Ministry of Education,College of Geographic Science,Nanjing Normal University,Key Laboratory of Virtual Geographic Environment,Ministry of Education,College of Geographic Science,Nanjing Normal University,Key Laboratory of Virtual Geographic Environment,Ministry of Education,College of Geographic Science,Nanjing Normal University,Key Laboratory of Virtual Geographic Environment,Ministry of Education,College of Geographic Science,Nanjing Normal University,Key Laboratory of Virtual Geographic Environment,Ministry of Education,College of Geographic Science,Nanjing Normal University,Key Laboratory of Virtual Geographic Environment,Ministry of Education,College of Geographic Science,Nanjing Normal University,Key Laboratory of Virtual Geographic Environment,Ministry of Education,College of Geographic Science,Nanjing Normal University,Key Laboratory of Virtual Geographic Environment,Ministry of Education,College of Geographic Science,Nanjing Normal University Fund Project: 摘要 | 图/表 | 访问统计 | 参考文献 | 相似文献 | 引证文献 | 资源附件 | 文章评论 摘要:悬浮物直接影响到光在水体中的传播,进而影响着水生生态环境,最终决定了湖泊的初级生产力。传统的遥感反演估算模型大多是针对某一湖区进行统一建模,忽视了不同区域水体光学性质的复杂差异性,并且传统的传感器时间分辨率和空间分辨率受到一定限制。针对太湖、巢湖、滇池、洞庭湖4个湖区利用两步聚类法将高光谱模拟到GOCI影像上的波段进行分类,将水体类型分为三类,第一类水体为悬浮物主导的水体,第二类水体为悬浮物和叶绿素a共同主导的水体,第三类水体为叶绿素a主导的水体。针对不同类型水体的光学特征,分别构建了悬浮物浓度反演模型,结果表明第一类水体可以利用B7/B4,第二和第三类水体可以利用B7/(B8+B4)作为波段组合因子对悬浮物浓度进行模型构建。精度验证结果表明,分类建模后第一类和第三类水体悬浮物浓度估算精度都得到了较明显提高,第一类水体RMSE降低了9.19mg/L,MAPE降低了3%,第三类水体RMSE降低了5.63 mg/L,MAPE降低了13.97%,第二类水体精度稍有降低。最后将反演模型应用于2013年5月13日的GOCI影像,可知整体而言太湖西南部地区悬浮物浓度较高,东北部地区悬浮物浓度较低,并且从9:00到15:00,太湖南部悬浮物浓度较高的区域在逐渐缩小。 Abstract:Total suspended matter (TSM) is an important water quality indicator that can directly affect the propagation of light in water and influence the aquatic ecological environment, and ultimately determines the primary productivity of a lake. Empirical TSM concentration estimation models are often built for specific study areas, ignoring variation in the optical properties of water among diverse areas. In addition, common satellite sensors cannot be successfully used to monitor inland lakes owing to their temporal and spatial resolution. Taihu Lake, Chaohu Lake, Dianchi Lake, and Dongting Lake were selected as our study lakes, and an automatic two-step cluster method was applied for water classification based on simulated geostationary ocean color imager (GOCI) reflectance spectra. The results showed that the water samples could be classified into three types. The optical features of Water Type 1 were influenced by the TSM, the optical characteristics of Water Type 2 were influenced by both TSM and chlorophyll-a (Chl-a), and the optical properties of Water Type 3 were mainly determined by Chl-a. Estimation models were then developed for each water type using a band ratio of B7/B4 for Water Type 1 and B7/(B8 + B4) for Water Types 2 and 3 to retrieve the concentration of suspended solids. The root mean-squared errors (RMSEs) and minimum absolute percentage errors (MAPEs) of Water Type 1 were 9.19 mg/L and 3%, and those of Water Type 3 were 5.63 mg/L and 13.97%, respectively, which were significantly lower than those estimated using methods that do not consider this classification. The RMSE and MAPE of Water Type 2 were slightly higher than those estimated with the general algorithm. The diurnal variation of the TSM concentration in Taihu Lake was studied based on the GOCI data acquired on May 13, 2013 using this classification method, and the results showed that the concentration of TSM was higher in the southwest than in the northeast. In addition, the area of higher TSM concentration in the southern region of the lake was reduced from 9:00 to 15:00(Beijing Local Time). 参考文献 相似文献 引证文献

Multiobjective Optimization of Multiloop Fractional Order PID Controller Tuned Using Bat Algorithm for Two Interacting Conical Tank Process

Multiloop fractional order PID controller is tuned using Bat algorithm for two interacting conical tank process. Two interacting conical tank process is modelled using mass balance equations. Two Interacting Conical Tank process is a complex system involving tedious interaction. Straight forward multiloop PID controller design involves various steps to design the controller. Due to easy implementation and quick convergence, Bat algorithm is used in recent past for solving continuous non-linear optimization problems to achieve global optimal solution. Bat algorithm, a swarm intelligence technique will be attempted to tune the multiloop fractional order PID controller for two interacting conical tank process. The multi objective optimized multiloop fractional PID controller is tested for tracking, disturbance rejection for minimum Integral time absolute error.

Inflation Forecast Based on BP Neural Network Model

This paper analyzes inflation forecast based on BP neural network model. Firstly, it reviews some references about BP neural network and finds that it is a nonlinear adaptive data-driven model with induction ability and a wide range of function approximation ability so that BP neural network could be applied into forecast research. Secondly, it builds up the BP neural network model to predict CPI, selecting the four indicators, which are excess liquidity, exchange rates, inflation expectation and macro-economic leading index. Then it carries out empirical experiment and takes advantage of the monthly data of the above four indicators from March 2005 to December 2012 to forecast CPI. The results show that when prediction period is 3 months, the maximum absolute error between forecast value and real value is 0.0139, and the minimum absolute error is 0.0005. When prediction period is 6 months, the maximum absolute error is not more than 0.02. It proves that BP neural network model can predict coming CPI trend at least 6 months according to the existing data and it means it is suitable for the study of inflation forecast.

Research of EDM Titanium Alloy TC11 Based on GM(1,N) Model and BP Neural Network

In order to predict the influences of electrical discharge machining (EDM) and enhance material removal rate (MRR). A model was presented for predictions of MRR in EDM TC11 test piece based on the construction method of gray GM(1,N) model and BP neural network model (GNNM).The results show that the GNNM is credible with the maximum absolute error 3.58%, the minimum absolute error 0.01% and the mean error 0.53%were obtained .The results indicated that the model can reflect the technological law of EDM TC11 and successfully predict its processing speed .The paper provides references and basis for selecting processing parameters of EDM TC11 and has practical significance.

Range Limited Quad-Histogram Equalization Method for Image Contrast Enhancement

equalization is a very popular technique for image contrast enhancement. Histogram equalization defines the clarity of picture and sometime improves brightness. But sometime it is not very good for consumer electronics, because of preserving the original brightness. Range limited bi-histogram equalization RLBHE applied for consumer electronics. This paper proposes novel extension of RLBHE referred to as range limited quad- histogram equalization RLQHE. This method is applied as first it divides the input histogram into two sub-images by a threshold. Then again two sub-images divide in to two sub- images by using another threshold that minimizes intra- class variance. All these sub-images separate the object from background. Then, Range means it calculate the minimum absolute mean brightness error between actual image and equalized image. The result shows that proposed method is better than previous method for image contrast enhancement. Proposed method preserves original brightness as actual image. It is very well suited for consumer electronics product.

Blind End-Member and Abundance Extraction for Multispectral Fluorescence Lifetime Imaging Microscopy Data

This paper proposes a new blind end-member and abundance extraction (BEAE) method for multispectral fluorescence lifetime imaging microscopy (m-FLIM) data. The chemometrical analysis relies on an iterative estimation of the fluorescence decay end-members and their abundances. The proposed method is based on a linear mixture model with positivity and sum-to-one restrictions on the abundances and end-members to compensate for signature variability. The synthesis procedure depends on a quadratic optimization problem, which is solved by an alternating least-squares structure over convex sets. The BEAE strategy only assumes that the number of components in the analyzed sample is known a spriori. The proposed method is first validated by using synthetic m-FLIM datasets at 15, 20, and 25dB signal-to-noise ratios. The samples simulate the mixed response of tissue containing multiple fluorescent intensity decays. Furthermore, the results were also validated with six m-FLIM datasets from fresh postmortem human coronary atherosclerotic plaques. A quantitative evaluation of the BEAE was made against two popular techniques: minimum volume constrained nonnegative matrix factorization (MVC-NMF) and multivariate curve resolution-alternating least-squares (MCR-ALS). Our proposed method (BEAE) was able to provide more accurate estimations of the end-members: 0.32% minimum relative error and 13.82% worst-case scenario, despite different initial conditions in the iterative optimization procedure and noise effect. Meanwhile, MVC-NMF and MCR-ALS presented more variability in estimating the end-members: 0.35% and 0.34% for minimum errors and 15.31% and 13.25% in the worst-case scenarios, respectively. This tendency was also maintained for the abundances, where BEAE obtained 0.05 as the minimum absolute error and 0.12 in the worst-case scenario; MCR-ALS and MVC-NMF achieved 0.04 and 0.06 for the minimum absolute errors, and 0.15 and 0.17 under the worst-case conditions, respectively. In addition, the average computation time was evaluated for the synthetic datasets, where MVC-NMF achieved the fastest time, followed by BEAE and finally MCR-ALS. Consequently, BEAE improved MVC-NMF in convergence to a local optimal solution and robustness against signal variability, and it is roughly 3.6 time faster than MCR-ALS.

An Analytical Model for Evaluating Static Power of Homogeneous FPGA Architectures

As capacity of the field-programmable gate arrays (FPGAs) continues to increase, power dissipated in the logic and routing resources has become a critical concern for FPGA architects. Recent studies have shown that static power is fast approaching the dynamic power in submicron devices. In this article, we propose an analytical model for relating homogeneous island-style-based FPGA architecture to static power. Current FPGA power models are tightly coupled with CAD tools. Our CAD-independent model captures the static power for a given FPGA architecture based on estimates of routing and logic resource utilizations from a pre-technology mapped netlist. We observe an average correlation ratio (C-Ratio) of 95% and a minimum absolute percentage error (MAPE) rate of 15% with respect to the experimental results generated by the Versatile Placement Routing (VPR) tool over the MCNC benchmarks. Our model offers application engineers and FPGA architects the capability to evaluate the impact of their design choices on static power without having to go through CAD-intensive investigations.

Tomotherapy treatment plan quality assurance: The impact of applied criteria on passing rate in gamma index method

Pretreatment patient plan verification with gamma index (GI) metric analysis is standard procedure for intensity modulated radiation therapy (IMRT) treatment. The aim of this paper is to evaluate the variability of the local and global gamma index obtained during standard pretreatment quality assurance (QA) measurements for plans performed with Tomotherapy unit. The QA measurements were performed with a 3D diode array, using variable passing criteria: 3%∕3 mm, 2%∕2 mm, 1%∕1 mm, each with both local and global normalization. The authors analyzed the pretreatment QA results for 73 verifications; 37 were prostate cancer plans, 16 were head and neck plans, and 20 were other clinical sites. All plans were treated using the Tomotherapy Hi-Art System. Pretreatment QA plans were performed with the commercially available 3D diode array ArcCHECK™. This device has 1386 diodes arranged in a helical geometry spaced 1 cm apart. The dose measurements were acquired on the ArcCHECK™ and then compared with the calculated dose using the standard gamma analysis method. The gamma passing rate (%GP), defined as the percentage of points satisfying the condition GI < 1, was calculated for different criteria (3%∕3 mm, 2%∕2 mm, 1%∕1 mm) and for both global and local normalization. In the case of local normalization method, the authors set three dose difference threshold (DDT) values of 2, 3, and 5 cGy. Dose difference threshold is defined as the minimum absolute dose error considered in the analysis when using local normalization. Low-dose thresholds (TH) of 5% and 10% were also applied and analyzed. Performing a paired-t-test, the authors determined that the gamma passing rate is independent of the threshold values for all of the adopted criteria (5%TH vs 10%TH, p > 0.1). Our findings showed that mean %GPs for local (or global) normalization for the entire study group were 93% (98%), 84% (92%), and 66% (61%) for 3%∕3 mm, 2%∕2 mm, and 1%∕1 mm criteria, respectively. DDT was equal to 2 cGy for the local normalization analysis cases. The authors observed great variability in the resulting %GP. With 3%∕3 mm gamma criteria, the overall passing rate with local normalization was 4.6% less on the average than with global one, as expected. The wide difference between %GP calculated with global or local approach is also confirmed by an unpaired t-test statistical analysis. The variability of %GP obtained confirmed the necessity to establish defined agreement criteria that could be universal and comparable between institutions. In particular, while the gamma passing rate does not depend on the choice of threshold, the choice of DDT strongly influences the gamma passing rate for local calculations. The difference between global and local %GP was statistically significant for prostate and other treatment sites when DDT was changed from 2 to 3 cGy.

Prediction of Rainfall in India using Artificial Neural Network (ANN) Models

In this paper, ARIMA(1,1,1) model and Artificial Neural Network (ANN) models like Multi Layer Perceptron (MLP), Functional-link Artificial Neural Network (FLANN) and Legendre Polynomial Equation ( LPE) were used to predict the time series data. MLP, FLANN and LPE gave very accurate results for complex time series model. All the Artificial Neural Network model results matched closely with the ARIMA(1,1,1) model with minimum Absolute Average Percentage Error(AAPE). Comparing the different ANN models for time series analysis, it was found that FLANN gives better prediction results as compared to ARIMA model with less Absolute Average Percentage Error (AAPE) for the measured rainfall data.

Development of the positive mean stress diagrams using genetic algorithm approach

ABSTRACTIn this numerical study, a new optimum positive mean stress fatigue failure equation is developed using previous experimental data and genetic algorithm. Two independent curve fitting coefficients are implemented in the equation to supply better correlations with experimental data of the failure strength envelope. In the literature, Gerber, Goodman, Soderberg, Morrow, Bagci, ASME (elliptic line), Clemson, Sekercioglu and so on suggested different equations for estimating fatigue strength envelope under mean stress condition. In these models, the effect of the materials was not considered in details. Some of these mean stress linear expressions are very conservative or have stress area bigger than the yield limit. The yield strength and the effect of materials are considered in the proposed model. The values of the positive mean stress, which correspond to fatigue failure, are obtained, and a minimum average absolute error among the models presented in the literature is remarked. The proposed model, which has less conservative structure, can be effectively used in the mechanical design process.

DETERMINATION OF ALL STABILIZING PI CONTROLLERS FOR LINEAR TIME INVARIANT SYSTEMS

This paper considers the problem of controlling a given linear time invariant (LTI) system described by a rational transfer function using a PI controller. A method based on graphical and computational tools has been proposed. This method specifies a PI controller (Kp and Ki ) that satisfies both frequency domain specifications (a specified gain or phase margin) and minimum integral time absolute error (ITAE).

Estimation of evapotranspiration and its partition based on an extended three-temperature model and MODIS products

Evapotranspiration (ET) is an energy balance component and a key component of water budget; thus, accurate estimates of ET are critical for understanding hydrological processes and water resources management. Despite growing concerns, challenges remain in estimating ET by remote sensing technology for regional applications, due to the difficulty in determination of aerodynamic, canopy and soil resistance. Instead of using resistance, ET and their partition of evaporation (Es) and transpiration (Ec) can be evaluated based on an extended three temperature model (3T model) and Moderate Resolution Imaging Spectroradiometer (MODIS) products for the period of 2001–2009. A case study was conducted in the Heihe River Basin in northwestern China. Validation results indicated that the mean absolute error was 0.08mm/d, with a maximum and minimum absolute error of 1.28mm/d and 0.02mm/d, respectively, which presented a simple extended 3T model in estimating ET with adequate accuracy and could satisfy regional research requirement at large scales. Results showed that: (1) seasonally, ET was highest (varied from 0.80mm/d to 1.27mm/d with a mean value of 1.05mm/d) in summer and lowest in winter (varied from 0.35mm/d to 0.44mm/d with a mean value of 0.37mm/d); (2) spatially, it rendered a decrease from the Qilian Mountain of the upper reaches (515mm/a) to the Gobi desert of the lower reaches (82mm/a), which was closely related to land covers and climate conditions; (3) the comparison of different ecosystems indicated that forest has the highest ET (588mm), followed by grassland (308mm), farmland (225mm), and desert land (160mm); (4) ET was separated into Es and Ec, and about 38–72% soil water returned to the atmosphere in the form of Es. This information will prove to be useful for water-use efficiency of the Ecological Water Conveyances Project (EWCP) conducted in the basin.

An Investigation of Using Different Saturation Height Functions by Introducing Rock Type Property

Estimation of the water saturation greatly impacts oil-in-place calculations. Saturation-height function is used to predict the saturation in the reservoir for a given height above the free water level. Different saturation-height functions have been presented so far, of which all of them are based upon using laboratory information to estimate water saturation. In this article, first the performance of nine saturation-height methods presented so far is investigated for the overall study of the reservoir. In the next stage, the reservoir has been divided into different rock types and water saturation estimated by different methods in each rock type classification. The results indicated that the Kamalyar and Leverett method with coefficient independent of core properties could be the best and worst options for overall study of the reservoir, respectively. By using the Winland method, the Kamalyar model indicated the minimum average absolute error among the other eight theoretical and empirical methods at both R35 values equal to 0.1–0.5 and 0.5–2 micron, while Sodena and Leverett methods have the highest prediction error for water saturation prediction, respectively. By using the discrete rock type method, for discrete rock type values equal to 8 and 9, the Johnson and Kamalyar models have the minimum average absolute error, respectively. In both discrete rock type values, the Leverett method gives the highest prediction error.

Precision pointing estimator design for minimum absolute, window- and stability-time errors

Future satellite missions have high precision pointing performance requirements. This leads to the necessity of optimizing attitude estimators not only with respect to absolute, but also with respect to window- and stability-time errors. These errors are especially important for achieving unprecedented image quality. The standard attitude estimator is designed to be optimal with respect to its absolute knowledge error. In practice, such an estimator design is either assumed to be almost optimal for window- and stability-time errors or a cumbersome simulation-based optimization is conducted to achieve optimality in this respect. The first assumption is not sufficient for high precision pointing missions and simulation-based design is computationally expensive. In this article frequency-domain metrics are used to develop a continuous-time design approach for a fixed gain attitude estimator that is optimized with respect to window- or stability-time errors. Colored noise is explicitly taken into account in the optimization because its spectrum strongly affects the solution. However, that might lead to instability of the estimator. This issue is addressed and a solution is provided that guarantees stability with the burden of marginal performance losses only. The approach is applied to design a two-state attitude estimator and its results are verified in frequency-domain computations and time-domain simulations for the future ESA mission, Meteosat Third Generation. In this case, the computational time for optimizing the window-time error has been reduced from hours to a few seconds. The fast and high precision estimator design for window- and stability-time errors thus supports time-efficient design trade-offs.

الذکاء الاصطناعی وتحلیل السلاسل الزمنیة

هذا البحث يعتبر امتدادا لبحث Alshawadfi (2003) والذي قدم فيه طريقة جديدة ومؤثرة للتنبؤ بنماذج ARMA, باستخدام أحد طرق الذکاء الاصطناعي وهي طريقة الشبکات العصبية, کما قارن بين الطريقة المقترحة وطريقة بوکس-جينکنز.ويتضمن هذا البحث هدفين: أولهما: تعميم طريقة Alshawadfi للتنبؤ بالسلاسل الزمنية المولدة من نماذج ARMAX باستخدام الشبکات العصبية الاصطناعية (ANN), ولتحقيق هذا الهدف تم توليد32000 عينة بأحجام مختلفة (25-40-60-100-150) من نماذج ARMAX بمعلمات مختلفة تستخدم لتدريب الشبکة,بعد ذلک تم مقارنة البيانات المولدة بالبيانات الحقيقية لقياس دقة التنبؤات. الهدف الثاني: مقارنة أداء الطريقة المقترحة للشبکات العصبية الاصطناعية مع أداء طريقة بوکس و جينکنز لتوضيح أي منهما أفضل للتنبؤ بنماذج ARMAX وذلک من خلال قياس دقة التنبؤات لکل من الطريقتين باستخدام ثلاث طرق هي متوسط مربعات الخطأ Mean of Squared Error (MSE), متوسط الانحرافات المطلقة Mean Absolute Deviation (MAD), نسبة الحالات التي تتحقق فيها أخطاء مطلقة أقل في الطريقتين The Percentage of cases of Minimum Absolute error (MAEP). ولانجاز عملية تدريب الشبکة واختبارها وحساب تنبؤات طريقتي بوکس-جينکنز والشبکات العصبية مع المقارنة بين الطريقتين تم استخدام حزم برامج ماتلاب MATLAB. وتم تصميم برنامج کمبيوتر مناسب ضمن حزم برامج ماتلاب MATLAB (أنظر الملحق المرفق مع البحث ) لتوليد البيانات و لانجاز عملية تدريب الشبکة واختبارها وحساب تنبؤات طريقتي بوکس-جينکنز والشبکات العصبية مع المقارنة بينهما. وقد أظهرت نتائج البحث وجود قدرة عالية للطريقة المقترحة للتنبؤ لنماذج ARMAX باستخدام أحد طرق الذکاء الاصطناعي - وهي طريقة الشبکات العصبية - على التنبؤ بالقيم المستقبلية للسلسلة الزمنية المعطاة وذلک بطريقة آلية, حيث أظهرت النتائج من دراسة 32000 عينة والتي تم توليدها أن متوسط مربعات أخطاء التنبؤMSE خاصة في حالة العينات الصغيرة, ومتوسط القيمة المطلقة لخطأ التنبؤ MAD, وکذلک متوسط النسبة MAEP أفضل من نظيرتها في أسلوب بوکس-جينکنز, وذلک کمتوسط عام بالنسبة لکل العينات وکل النماذج وللمشاهدات المستقبلية الثلاث الأولى. وبناءا على ذلک يمکن القول بصلاحية الطريقة المقترحة للتنبؤ بالقيم المستقبلية للسلاسل الزمنية المولدة من نماذج ARMAX. This paper has two objects. First, we present artificial neural networks method for forecasting linear and nonlinear ARAMAX time series. Second, we compare the proposed method with the well known Box-Jenkins method through a simulation study . To achieve these objects 32000 samples, generated from different ARMAX models, different sizes (25,40,60,100,150), were used for the network training. Then the system was tested for generated data . The accuracy of the neural network forecasts(NNF) is compared with the corresponding Box-Jenkins forecasts(BJF) by using three tools: the mean square error (MSE) , the mean absolute deviation of error (MAD) and the ratio of closeness from the true values (MPE) . A suitable computer program was designed (MATLAB TOOLBOX) for NN training , testing and comparing with Box-Jenkins method . The forecasts of theproposed NN approach, as shown from three measures, seem to provide better results than the classical forecasting Box-Jenkins approach . The results suggest thatthe ANN approach may provide a superior alternative to the Box-Jenkins forecasting approach for developing forecasting models in situations that do not require modeling of the internal structure of the series . The numerical results show that the proposed approach has a good performance for the forecasting of ARMAX(p,h,q) models.

An investigation into direct fabrication of fine-structured components by selective laser melting

This paper presents the direct fabrication of 3 mm-sized components applied in some practical fields using a selective laser melting (SLM) machine with fiber laser: a radiator with variable-section thin wall, an optical component with required smooth surface, and a throttle with a complicated cavity. Dimensional precision, geometric defect, and selection of fabrication direction were discussed after the implementing experiments. Experimental results showed that the minimum absolute error in the wall thickness of the radiator was about 20 μm, the roughness of the optical component should be improved by post-processes, and the cavity of the throttle could be successfully fabricated using tiling display. The current study raises several issues during selective laser melting of fine-structured components which need to be solved before expansion of SLM to practical applications.