行人惯性导航零速检测算法

在行人惯性导航中,零速检测是实现速度误差清零和导航误差估计的前提,有着重要的作用.针对行人运动过程中零速区间时间间隔短难以检测的问题,提出了一种基于人体脚部运动特征的零速检测算法,将步行运动抽象成了一个包含4个隐含状态与15个观测量的隐马尔可夫模型,并阐述了模型构建机理.利用Baum-Welch算法训练和优化模型参数,提高了检测准确率.实验结果表明:所提出的方法零速检测效果较好,且采用该方法的行人惯性导航系统,其定位误差约为行进距离的0.73％,定位精度较高.     

行人导航系统中航迹推算参数估计方法的研究

将GPS/DR组合导航技术应用到行人导航系统中。利用加速度计对行人步态进行判别,将神经网络用于行人步幅信息的标定。同时利用电子罗盘实现方位角的测量,并结合加速度计信息进行倾斜角度的误差补偿。现场测试结果表明,本文提出的DR参数估计方法不但提高参数估计的精度,而且能够满足行人导航定位的要求。     

Application of Heuristic Asymmetric Mapping for mobile robot navigation using ultrasonic sensors

In this paper, an experimental study of a navigation system that allows a mobile robot to travel in an environment about which it has no prior knowledge is described. Data from multiple ultrasonic range sensors are fused into a representation called Heuristic Asymmetric Mapping to deal with the problem of uncertainties in the raw sensory data caused mainly by the transducer's beam-opening angle and specular reflections. It features a fast data-refresh rate to handle a dynamic environment. Potential-field method is used for on-line path planning based on the constructed gridtype sonar map. The mobile robot can therefore learn to find a safe path according to its self-built sonar map. To solve the problem of local minima in conventional potential field method, a new type of potential function is formulated. This new method is simple and fast in execution using the concept from distance-transform path-finding algorithms. The developed navigation system has been tested on our experimental mobile robot to demonstrate its possible application in practical situations. Several interesting simulation and experimental results are presented.This work was supported partly by the National Science Council of Taiwan, ROC under the grant NSC-82-0422-E-009-321.     

A novel multiplex cascade classifier for pedestrian detection

Reliable pedestrian detection is of great importance in visual surveillance. In this paper, we propose a novel multiplex classifier model, which is composed of two multiplex cascades parts: Haar-like cascade classifier and shapelet cascade classifier. The Haar-like cascade classifier filters out most of irrelevant image background, while the shapelet cascade classifier detects intensively head-shoulder features. The weighted linear regression model is introduced to train its weak classifiers. We also introduce a structure table to label the foreground pixels by means of background differences. The experimental results illustrate that our classifier model provides satisfying detection accuracy. In particular, our detection approach can also perform well for low resolution and relatively complicated backgrounds.     

Obstacle avoidance algorithm for visual navigation using ultrasonic sensors and a CCD camera

An obstacle-avoidance algorithm is presented for autonomous mobile robots equipped with a CCD camera and ultrasonic sensors. This approach uses segmentation techniques to segregate the floor from other fixtures, and measurement techniques to measure the distance between the mobile robot and any obstacles. It uses a simple computation for the selection of a threshold value. This approach also uses a cost function, which is combined with image information, distance information, and a weight factor, to find an obstacle-free path. This algorithm, which uses a CCD camera and ultrasonic sensors, can be used for cases including shadow regions, and obstacles in visual navigation and in various lighting conditions. This work was presented in part at the 7th International Symposium on Artificial Life and Robotics, Oita, Japan, January 16–18, 2002     

A purely reactive navigation scheme for dynamic environments using Case-Based Reasoning

This paper presents a new sonar based purely reactive navigation technique for mobile platforms. The method relies on Case-Based Reasoning to adapt itself to any robot and environment through learning, both by observation and self experience. Thus, unlike in other reactive techniques, kinematics or dynamics do not need to be explicitly taken into account. Also, learning from different sources allows combination of their advantages into a safe and smooth path to the goal. The method has been succesfully implemented on a Pioneer robot wielding 8 Polaroid sonar sensors. Cristina Urdiales is a Lecturer at the Department of Tecnología Electrónica (DTE) of the University of Málaga (UMA). She received a MSc degree in Telecommunication Engineering at the Universidad Politécnica de Madrid (UPM) and her Ph.D. degree at University of Málaga (UMA). Her research is focused on robotics and computer vision. E.J. Pérez was born in Barcelona, Spain, in 1974. He received his title of Telecommunication Engineering from the University of Málaga, Spain, in 1999. During 1999 he worked in a research project under a grant by the Spanish CYCIT. From 2000 to the present day he has worked as Assistant Professor in the Department of Tecnología Electrónica of the University of Málaga. His research is focused on robotics and artificial vision. Javier Vázquez-Salceda is an Associate Researcher of the Artificial Intelligence Section of the Software Department (LSI), at the Technical University of Catalonia (UPC). Javier obtained an MSc degree in Computer Science at UPC. After his master studies he became research assistant in the KEMLg Group at UPC. In 2003 he presented his Ph.D. dissertation (with honours), which has been awarded with the 2003 ECCAI Artificial Intelligence Dissertation Award. The dissertation has been also recently published as a book by Birkhauser-Verlag. From 2003 to 2005 he was researcher in the Intelligent Systems Group at Utrecht University. Currently he is again member of the KEMLg Group at UPC. His research is focused on theoretical and applied issues of Normative Systems, software and physical agents' autonomy and social control, especially in distributed applications for complex domains such as eCommerce or Medicine. Miquel Sànchez-Marrè (Barcelona, 1964) received a Ph.D. in Computer Science in 1996 from the Technical University of Catalonia (UPC). He is Associate Professor in the Computer Software Department (LSI) of the UPC since 1990 (tenure 1996). He was the head of the Artificial Intelligence section of LSI (1997–2000). He is a pioneer member of International Environmental Modelling and Software Society (IEMSS) and a board member of IEMSS also, since 2000. He is a member of the Editorial Board of International Journal of Applied Intelligence, since October 2001. Since October 2004 he is Associate Editor of Environmental Modelling and Software journal. His main research topics are case-based reasoning, machine learning, knowledge acquisition and data mining, knowledge engineering, intelligent decision-support systems, and integrated AI architectures. He has an special interest on the application of AI techniques to Environmental Decision Support Systems. Francisco Sandoval was born in Spain in 1947. He received the title of Telecommunication Engineering and Ph.D. degree from the Technical University of Madrid, Spain, in 1972 and 1980, respectively. From 1972 to 1989 he was engaged in teaching and research in the fields of opto-electronics and integrated circuits in the Universidad Politécnica de Madrid (UPM) as an Assistant Professor and a Lecturer successively. In 1990 he joined the University of Málaga as Full Professor in the Department of Tecnología Electrónica. He is currently involved in autonomous systems and foveal vision, application of Artificial Neural Networks to Energy Management Systems, and in Broad Band and Multimedia Communication.     

Nonlinear regression in environmental sciences using extreme learning machines: A comparative evaluation

The extreme learning machine (ELM), a single-hidden layer feedforward neural network algorithm, was tested on nine environmental regression problems. The prediction accuracy and computational speed of the ensemble ELM were evaluated against multiple linear regression (MLR) and three nonlinear machine learning (ML) techniques – artificial neural network (ANN), support vector regression and random forest (RF). Simple automated algorithms were used to estimate the parameters (e.g. number of hidden neurons) needed for model training. Scaling the range of the random weights in ELM improved its performance. Excluding large datasets (with large number of cases and predictors), ELM tended to be the fastest among the nonlinear models. For large datasets, RF tended to be the fastest. ANN and ELM had similar skills, but ELM was much faster than ANN except for large datasets. Generally, the tested ML techniques outperformed MLR, but no single method was best for all the nine datasets.     

A framework using cluster-based hybrid network architecture for collaborative virtual surgery

Research on collaborative virtual environments (CVEs) opens the opportunity for simulating the cooperative work in surgical operations. It is however a challenging task to implement a high performance collaborative surgical simulation system because of the difficulty in maintaining state consistency with minimum network latencies, especially when sophisticated deformable models and haptics are involved. In this paper, an integrated framework using cluster-based hybrid network architecture is proposed to support collaborative virtual surgery. Multicast transmission is employed to transmit updated information among participants in order to reduce network latencies, while system consistency is maintained by an administrative server. Reliable multicast is implemented using distributed message acknowledgment based on cluster cooperation and sliding window technique. The robustness of the framework is guaranteed by the failure detection chain which enables smooth transition when participants join and leave the collaboration, including normal and involuntary leaving. Communication overhead is further reduced by implementing a number of management approaches such as computational policies and collaborative mechanisms. The feasibility of the proposed framework is demonstrated by successfully extending an existing standalone orthopedic surgery trainer into a collaborative simulation system. A series of experiments have been conducted to evaluate the system performance. The results demonstrate that the proposed framework is capable of supporting collaborative surgical simulation.     

A simulation tool for managing environmental flows in regulated rivers

Environmental flows provide river flow regimes to restore and conserve aquatic ecosystems, creating considerably different demands compared to conventional water extraction. With increasing incorporation of environmental flows in water planning worldwide, governments require decision support tools to manage these flows in regulated rivers. We developed the Environmental Water Allocation Simulator with Hydrology (eWASH), a fast, flexible and user-friendly scenario-based hydrological modelling tool, supporting environmental flow management decisions for single- or multi-reservoir systems. Environmental flow demands and management rules are easily specified via the graphical user interface, and batch processing functions aid in uncertainty assessment. eWASH modelled main processes of complex regulated rivers and the tool is widely applicable. We calibrated eWASH for the Gwydir and Macquarie Rivers of Australia's Murray–Darling Basin. Modelled monthly environmental flow allocations exhibited Nash–Sutcliffe efficiencies of 0.55 for the Gwydir and 0.72 for the Macquarie catchments respectively when validated.     

A line feature based SLAM with low grade range sensors using geometric constraints and active exploration for mobile robot

This paper describes a geometrically constrained Extended Kalman Filter (EKF) framework for a line feature based SLAM, which is applicable to a rectangular indoor environment. Its focus is on how to handle sparse and noisy sensor data, such as PSD infrared sensors with limited range and limited number, in order to develop a low-cost navigation system. It has been applied to a vacuum cleaning robot in our research. In order to meet the real-time objective with low computing power, we develop an efficient line feature extraction algorithm based upon an iterative end point fit (IEPF) technique assisted by our constrained version of the Hough transform. It uses a geometric constraint that every line is orthogonal or parallel to each other because in a general indoor setting, most furniture and walls satisfy this constraint. By adding this constraint to the measurement model of EKF, we build a geometrically constrained EKF framework which can estimate line feature positions more accurately as well as allow their covariance matrices to converge more rapidly when compared to the case of an unconstrained EKF. The experimental results demonstrate the accuracy and robustness to the presence of sensor noise and errors in an actual indoor environment.

A modeling and spatio-temporal analysis framework for monitoring environmental change using NPP as an ecosystem indicator

We present and describe a modeling and analysis framework for monitoring protected area (PA) ecosystems with net primary productivity (NPP) as an indicator of health. It brings together satellite data, an ecosystem simulation model (NASA-CASA), spatial linear models with autoregression, and a GIS to provide practitioners a low-cost, accessible ecosystem monitoring and analysis system (EMAS) at landscape resolutions. The EMAS is evaluated and assessed with an application example in Yellowstone National Park aimed at identifying the causes and consequences of drought. Utilizing five predictor covariates (solar radiation, burn severity, soil productivity, temperature, and precipitation), spatio-temporal analysis revealed how landscape controls and climate (summer vegetation moisture stress) affected patterns of NPP according to vegetation functional type, species cover type, and successional stage. These results supported regional and national trends of NPP in relation to carbon fluxes and lag effects of climate. Overall, the EMAS provides valuable decision support for PAs regarding informed land use planning, conservation programs, vital sign monitoring, control programs (fire fuels, invasives, etc.), and restoration efforts.     

Integrating spectral indices with environmental parameters for estimating heavy metal concentrations in rice using a dynamic fuzzy neural-network model

A generalized dynamic fuzzy neural network (GDFNN) was created to estimate heavy metal concentrations in rice by integrating spectral indices and environmental parameters. Hyperspectral data, environmental parameters, and heavy metal content were collected from field experiments with different levels of heavy metal pollution (Cu and Cd). Input variables used in the GDFNN model were derived from 10 variables acquired by gray relational analysis. The assessment models for Cd and Cu concentration employed five and six input variables, respectively. The results showed that the GDFNN for estimating Cu and Cd concentrations in rice performed well at prediction with a compact network structure using the training, validation, and testing sets (for Cu, fuzzy rules=9, R² greater than 0.75, and RMSE less than 2.5; for Cd, fuzzy rules=9, R² greater than 0.75, and RMSE less than 1.0). The final GDFNN model was then compared with a back-propagation (BP) neural network model, adaptive-network-based fuzzy interference systems (ANFIS), and a regression model. The accuracies of GDFNN model prediction were usually slightly better than those of the other three models. This demonstrates that the GDFNN model is more suitable for predicting heavy metal concentrations in rice.     

Creation of environmental health information system for public health service: A pilot study

Despite more than a decade of research on medical information systems, deficiencies exist in our capability of establishing an effective environmental health information infrastructure. In this research, we present a pilot study on creating a feasible environmental health information infrastructure. The newly-developed environmental health information system is a web-based platform that integrates databases, decision-making tools, geographic information systems for supporting public health service and policy making. The study, which is a part of a comprehensive effort known as Environmental Public Health Tracking proposed by the Center for Disease Control and Prevention, opens the door for future research on a large scale nation-wide healthcare information infrastructure.

A fully neural implementation of unitary response model for classification of gases/odors using the responses of thick film gas sensor array

In this paper, a classification scheme based on neurally implemented unitary response model (URM) for a gas/odor sensor array response has been presented. Thick-film tin-oxide sensor array responses for four gases/odors (viz. acetone, carbon tetra-chloride, ethyl methyl ketone and xylene) were first transformed into equivalent unitary responses. This transformation was carried out using a pre-trained neural ‘unitary response model pre-processor (URMP)’, called Net I_URMP. The classification of these responses in the unitary analysis space was then carried out, more accurately, using a pre-trained neural classifier called Net II_URMC. During this experiment, respective nets Net I_URMP and Net II_URMC, comprising of 12 and 8 neurons, were trained in just 23 and 09 epochs of 42 × 4 training response vectors. At stage I, the mean squared error (MSE) between neurally and mathematically obtained unitary response versions of 18 independent test responses for the considered gases/odors was 7.51 × 10⁻². At stage II, all the aforesaid test samples were correctly classified, with a MSE of 3.87 × 10⁻⁸. Further, by connecting Net I_URMP and Net II_URMC in cascade, the proposed classifier could be implemented using 16 neurons only.