Advanced Artificial Neural Network Research Articles

New Trends for the Nonlinear Measurement and Modeling of High-Power RF Transistors and Amplifiers With Memory Effects

Power amplifier (PA) behavior is inextricably linked to the characteristics of the transistors underlying the PA design. All transistors exhibit some degree of memory effects, which must therefore be taken into account in the modeling and design of these PAs. In this paper, we will present new trends for the characterization, device modeling, and behavioral modeling of power transistors and amplifiers with strong memory effects. First the impact of thermal and electrical memory effects upon the performance of a transistor will be revealed by comparing continuous wave and pulsed RF large-signal measurements. Pulsed-RF load-pull from the proper hot bias condition yields a more realistic representation of the peak power response of transistors excited with modulated signals with high peak-to-average power ratio. Next, an advanced device modeling method based on large-signal data from a modern nonlinear vector network analyzer instrument, coupled with modeling approaches based on advanced artificial neural network technology, will be presented. This approach enables the generation of accurate and robust time-domain nonlinear simulation models of modern transistors that exhibit significant memory effects. Finally an extension of the X-parameter (X-parameter is a trademark of Agilent Technologies Inc.) behavioral model to account for model memory effects of RF and microwave components will be presented. The approach can be used to model hard nonlinear behavior and long-term memory effects and is valid for all possible modulation formats for all possible peak-to-average ratios and for a wide range of modulation bandwidths. Both the device and behavioral models have been validated by measurements and are implemented in a commercial nonlinear circuit simulator.

Indoor thermal condition in urban heat Island – Development of a predictive tool

Urban Heat Island (UHI) effects have caused extensive economic and health related issues to many city residents, especially the most vulnerable such as elderly people living in buildings without air conditioners or mechanical ventilation systems. To reinforce the resiliency of individuals and communities in facing extreme heat event, cities are developing reliable tools to predict the indoor thermal characteristics using available building characteristics, climate data and socio-economical factors.In this study, a novel approach is proposed to predict the indoor thermal conditions in these buildings. First, a measurement campaign is conducted to monitor indoor thermal condition within 55 buildings in most vulnerable regions on the Island of Montreal. Two models, Simplified and Advanced, are developed to predict hourly indoor dry-bulb temperatures. Both models use an advanced Artificial Neural Network (ANN) technique. The Simplified ANN Model generates a correlation between airport weather observations and monitored indoor dry-bulb temperatures. On the other hand, the Advanced Model includes ten influential parameters, which represent the effect of neighboring environment, building characteristics and its usage patterns on the indoor thermal condition. Comparison of these two predictive models is conducted on different levels of simulation and validation. The Advanced Model shows better accuracy in predicting the indoor thermal conditions, thus justifying the use of neighborhood specific parameters to forecast indoor environment condition in an urban heat island area.

Optimization of C.I Engine Parameters Using Artificial Neural

Optimization of Compression Ignition Engines through advanced artificial neural network is the modern process in mechanization and best utilization of modern technology for better economic scenarios in coming generation. This project deals with the feasibility of using artificial neural networks in combination with genetic algorithms to optimize the diesel engine settings. The engine is operated by using diesel and sunflower oil blends and the output parameters are calculated theoretically with the standard mechanical formulae and those manual experimental calculated values are used for training several neural networks with different various hidden layer [ n x m ] matrix combinations. The output values given by these trained networks are compared with experimental values and out of which the trained error values are taken for all networks.

Regional forest-fire susceptibility analysis in central Portugal using a probabilistic ratings procedure and artificial neural network weights assignment

Geographic information system analysis and artificial neural network modelling were combined to evaluate forest-fire susceptibility in the Central Portugal administrative area. Data on forest fire events, indicated by burnt areas during the years from 1990 to 2007, were identified from official records. Topographic, supporting infrastructures, vegetation cover, climatic, demographic and satellite-image data were collected, processed and integrated into a spatial database using geographic information system techniques. Eight fire-related factors were extracted from the collected data, including topographic slope and aspect, road density, viewsheds from fire watchtowers, land cover, Landsat Normalised Difference Vegetation Index, precipitation and population density. Ratings were calculated for the classes or categories of each factor using a frequency-probabilistic procedure. The thematic layers (burnt areas and fire-related factors) were analysed using an advanced artificial neural network model to calculate the relative weight of each factor in explaining the distribution of burnt areas. A forest-fire susceptibility index was calculated using the trained back-propagation artificial neural network weights and the frequency-probabilistic ratings, and then a general forest-fire susceptibility index map was constructed in geographic information system. Burnt areas were used to evaluate the forest-fire susceptibility index map, and the results showed an agreement of 78%. This forest-fire susceptibility map can be used in strategic and operational forest-fire management planning at the regional scale.

Approaches for delineating landslide hazard areas using different training sites in an advanced artificial neural network model

The current paper presents landslide hazard analysis around the Cameron area, Malaysia, using advanced artificial neural networks with the help of Geographic Information System (GIS) and remote sensing techniques. Landslide locations were determined in the study area by interpretation of aerial photographs and from field investigations. Topographical and geological data as well as satellite images were collected, processed, and constructed into a spatial database using GIS and image processing. Ten factors were selected for landslide hazard including: 1) factors related to topography as slope, aspect, and curvature; 2) factors related to geology as lithology and distance from lineament; 3) factors related to drainage as distance from drainage; and 4) factors extracted from TM satellite images as land cover and the vegetation index value. An advanced artificial neural network model has been used to analyze these factors in order to establish the landslide hazard map. The back-propagation training method has been used for the selection of the five different random training sites in order to calculate the factor’s weight and then the landslide hazard indices were computed for each of the five hazard maps. Finally, the landslide hazard maps (five cases) were prepared using GIS tools. Results of the landslides hazard maps have been verified using landslide test locations that were not used during the training phase of the neural network. Our findings of verification results show an accuracy of 69%, 75%, 70%, 83% and 86% for training sites 1, 2, 3, 4 and 5 respectively. GIS data was used to efficiently analyze the large volume of data, and the artificial neural network proved to be an effective tool for landslide hazard analysis. The verification results showed sufficient agreement between the presumptive hazard map and the existing data on landslide areas.

Regional landslide susceptibility analysis using back-propagation neural network model at Cameron Highland, Malaysia

This paper presents landslide susceptibility analysis around the Cameron Highlands area, Malaysia using a geographic information system (GIS) and remote sensing techniques. Landslide locations were identified in the study area from interpretation of aerial photographs and field surveys. Topographical, geological data and satellite images were collected, processed, and constructed into a spatial database using GIS and image processing. Ten landslide occurrence factors were selected as: topographic slope, topographic aspect, topographic curvature and distance from drainage, lithology and distance from lineament, soil type, rainfall, land cover from SPOT 5 satellite images, and the vegetation index value from SPOT 5 satellite image. These factors were analyzed using an advanced artificial neural network model to generate the landslide susceptibility map. Each factor’s weight was determined by the back-propagation training method. Then, the landslide susceptibility indices were calculated using the trained back-propagation weights, and finally, the landslide susceptibility map was generated using GIS tools. The results of the neural network model suggest that the effect of topographic slope has the highest weight value (0.205) which has more than two times among the other factors, followed by the distance from drainage (0.141) and then lithology (0.117). Landslide locations were used to validate the results of the landslide susceptibility map, and the verification results showed 83% accuracy. The validation results showed sufficient agreement between the computed susceptibility map and the existing data on landslide areas.

Classification of Erodible Soils Using Artificial Neural Networks and Biogeochemical Indicators

Past research has highlighted the need to classify non-point–pollution sources in a watershed in order to work toward environmental remediation of the streams and rivers. The present work couples field measurements of carbon and nitrogen isotopes and the carbon to nitrogen atomic ratio with artificial neural network, multivariate logistic regression, and multivariate discriminant analysis modeling to perform the classification of erodible soils from forest and agricultural land uses. Results show the importance of all three biogeochemical indicators in aiding classification and it is proposed that all three indicators be measured and evaluated in future studies. Multivariate logistic regression performed adequately in classification and at the same time represented the most parsimonious nonnormal model, which highlights the need to avoid creating nonparsimonious models by application of advanced artificial neural network modeling techniques without underlying rationale. Model misclassification occurs due to a shift in biogeochemical cycles in a subset of the agricultural erodible soil data that is reflective of conversion of soil from tilled winter-wheat/barley rotation to conservation, and misclassification of forest soils is attributed to the high variability of soil organic matter cycling. Quantification of samples that can be tagged as potentially misclassified is performed in order that these samples can be further analyzed prior to watershed remediation.

Utilization of Optical Remote Sensing Data and GIS Tools for Regional Landslide Hazard Analysis Using an Artificial Neural Network Model

The aim of this study is to evaluate landslide hazard analysis at Selangor area, Malaysia using optical remote sensing data and a Geographic Information System (GIS). Landslide locations were identified in the study area from interpretation of aerial photographs and field surveys. Topographical, geological data and satellite images were collected, processed and constructed into a spatial database using GIS and image processing. A total of 10 landslide occurrence factors that were selected including topographic slope, topographic aspect, topographic curvature and distance from drainage; lithology and distance from lineament; land cover from TM satellite images; the vegetation index value from Landsat satellite images; precipitation data. These factors were analyzed using an advanced artificial neural network model to generate the landslide hazard map. Each factor's weight was determined by the back-propagation training method. Then the landslide hazard indices were calculated using the trained back-propagation weights, and finally the landslide hazard map was generated using GIS tools. Landslide locations were used to verify results of the landslide hazard map and the verification results showed 82.92% accuracy. The verification results showed sufficient agreement between the presumptive hazard map and the existing data on landslide areas.

An artificial neural network to predict the outcome of repeat prostate biopsies

ObjectivesTo develop an advanced artificial neural network (ANN) to predict the presence of prostate cancer (PCa) and to predict the outcome of repeat prostate biopsies. The predictive accuracy was compared with the accuracy obtained using standard cutoffs for the free/total (f/t) prostate-specific antigen (PSA) ratio, PSA density (PSAD), PSA density of the transition zone (PSA-TZ), and the total and transition zone volumes. Clinical and biochemical diagnostic tests have been shown to improve PCa detection. When these tests are combined using an ANN, significant increases in specificity at high sensitivity are observed. MethodsThe Vienna-based multicenter European referral database for early PCa detection of 820 men with a PSA level between 4 and 10 ng/mL was used. The presence of PCa was determined using transrectal ultrasound-guided octant needle repeat biopsy. Variables in the database consisted of age, PSA, f/t PSA ratio, digital rectal examination findings, PSA velocity, and the transrectal ultrasound-guided variables of prostate volume, transition zone volume, PSAD, and PSA-TZ. The ANN used in the analysis was an advanced multilayer perceptron selected for accuracy by a genetic algorithm. ResultsThe repeat biopsy PCa detection rate was 10% (n = 83). At 95% sensitivity, the specificity for ANN was 68% compared with 54%, 33.5%, 21.4%, 14.7%, and 8.3% for multivariate logistic regression analysis, f/t PSA ratio, PSA-TZ, PSAD, and total PSA, respectively. The ANN reduced unnecessary repeat biopsies by 68% in this study. The area under the curve was 83% for the ANN versus 79%, 74.5%, 69.1%, 61.8%, and 60.5% for multivariate analysis, f/t PSA ratio, PSA-TZ, PSAD, and total PSA, respectively. ConclusionsThe current ANN found a strong pattern predictive of PCa in patients with a negative initial biopsy. By combining the individual clinical and biochemical markers into the ANN, 68% specificity at 95% sensitivity was achieved. The ANN allows more accurate and individual counseling of patients with a negative initial biopsy.

Artificial neural networks for modelling dynamics of impacting bodies and vehicles

A new approach is demonstrated and tested for modelling the non-linear multi-body dynamics problem of two impacting bodies. An advanced recurrent artificial neural network (ANN) is used to model this highly non-linear dynamic problem. The equations of motion of the dynamic system are used to explain the required inputs for the ANN. The developed networks are trained to predict the acceleration profiles of the impacting objects. Each object is represented by a corresponding ANN. The mutual effect between the two objects, during the impact, is also considered by training the two ANNs simultaneously. Two numerical examples are used to demonstrate the ability of the new developed approach. The first is the dynamic impact problem of two non-linear lumped systems. The second is two different vehicles in frontal impacts. For each case, the problem is numerically solved for several impact scenarios and the results are used to prepare the data required for training and testing the developed ANNs. The results of the two numerical examples demonstrated the ability of the ANNs to store the non-linear dynamic characteristics of the impacting objects after a successful supervised training. The trained networks are capable of modelling non-linear multi-body dynamics and impact problems.

Vehicle crash modelling using recurrent neural networks

The initial velocity and structural characteristics of any vehicle are the main factors affecting the vehicle response in case of frontal impact. Finite Element (FE) simulations are essential tools for crashworthiness analysis, however, the FE models are getting bigger, which increases the simulation time and cost. In the current research, an advanced Artificial Neural Network (ANN) was used to store the nonlinear dynamic characteristics of the vehicle structure. Therefore, several impact scenarios can be analyzed quickly with much less computational cost by using the trained networks. The equation of motion of the dynamic system was used to define the inputs and outputs of the ANN. The system dynamics was included in the network performance and the recurrent back-propagation learning rule was adapted in training the network. The results of the numerical examples indicated that the recurrent ANN can accurately capture the frontal crash characteristics of the impacting structures, and predict the crash performance of the same structures for any other crash scenario within the training limits.