Novel Hybrid of LS-SVM and Kalman Filter for GPS/INS Integration

2010 ◽  
Vol 63 (2) ◽  
pp. 289-299 ◽  
Author(s):  
Zhenkai Xu ◽  
Yong Li ◽  
Chris Rizos ◽  
Xiaosu Xu
Keyword(s):  
Support Vector Machine ◽  
Kalman Filter ◽  
Inertial Sensors ◽  
Inertial Sensor ◽  
Micro Electro Mechanical System ◽  
Research Area ◽  
Optimal Number ◽  
Integrated System ◽  
Support Vector ◽  
Highly Nonlinear

Integration of Global Positioning System (GPS) and Inertial Navigation System (INS) technologies can overcome the drawbacks of the individual systems. One of the advantages is that the integrated solution can provide continuous navigation capability even during GPS outages. However, bridging the GPS outages is still a challenge when Micro-Electro-Mechanical System (MEMS) inertial sensors are used. Methods being currently explored by the research community include applying vehicle motion constraints, optimal smoother, and artificial intelligence (AI) techniques. In the research area of AI, the neural network (NN) approach has been extensively utilised up to the present. In an NN-based integrated system, a Kalman filter (KF) estimates position, velocity and attitude errors, as well as the inertial sensor errors, to output navigation solutions while GPS signals are available. At the same time, an NN is trained to map the vehicle dynamics with corresponding KF states, and to correct INS measurements when GPS measurements are unavailable. To achieve good performance it is critical to select suitable quality and an optimal number of samples for the NN. This is sometimes too rigorous a requirement which limits real world application of NN-based methods.The support vector machine (SVM) approach is based on the structural risk minimisation principle, instead of the minimised empirical error principle that is commonly implemented in an NN. The SVM can avoid local minimisation and over-fitting problems in an NN, and therefore potentially can achieve a higher level of global performance. This paper focuses on the least squares support vector machine (LS-SVM), which can solve highly nonlinear and noisy black-box modelling problems. This paper explores the application of the LS-SVM to aid the GPS/INS integrated system, especially during GPS outages. The paper describes the principles of the LS-SVM and of the KF hybrid method, and introduces the LS-SVM regression algorithm. Field test data is processed to evaluate the performance of the proposed approach.

scholarly journals Low-Cost MEMS-Based NARX Model for GPS-Denied Areas

2020 ◽  
pp. 58-70
Keyword(s):  
Inertial Sensors ◽  
Inertial Sensor ◽  
Low Cost ◽  
Autonomous Vehicle ◽  
Field Tests ◽  
Micro Electro Mechanical System ◽  
Integrated System ◽  
Main Challenge ◽  
Autonomous Vehicle Navigation ◽  
Narx Model

Autonomous vehicle navigation has witnessed a huge revolutionary revision regarding development in Micro-Electro Mechanical System (MEMS) technology. Most recently, Strapdown Inertial Navigation System (SDINS) has successfully been integrated with Global Positioning System (GPS). However, different grades of MEMS inertial sensors are available and choosing the convenient grade is quite important. Noises in inertial sensor are mostly treated through de-noising the additive errors to improve the precision of SDINS output. Unfortunately, integration in SDINS mechanization causes a growing in SDINS error output which considered the main challenge in integrating MEMS inertial sensors with GPS. This paper aims to promote the long-term performance of the MEMS-SDINS/GPS integrated system. A new integrated structure is proposed to model the nonlinearities that exist in SDINS dynamics in addition to the error uncertainty in the inertial sensors’ measurements. A robust Nonlinear AutoRegressive models with eXogenous inputs (NARX) based algorithm are designed for data fusion in the proposed GPS/INS integrated system. Validation for the proposed integrated system has been carried out using different field tests data in order to assess the accuracy of the system during GPS denied environment. The results obtained demonstrate that the proposed NARX model is applicative and satisfactory which shows a desired prediction performance.

scholarly journals INS/WSN-Integrated Navigation Utilizing LS-SVM andH∞Filtering

2012 ◽  
Vol 2012 ◽  
pp. 1-19 ◽  
Author(s):  
Yuan Xu ◽  
Xiyuan Chen ◽  
Qinghua Li
Keyword(s):  
Support Vector Machine ◽  
Kalman Filter ◽  
Back Propagation ◽  
Sampling Period ◽  
Integrated System ◽  
Support Vector ◽  
Integrated Navigation ◽  
New Approach ◽  
Stochastic Uncertainty ◽  
Integrated Navigation System

In order to achieve continuous navigation capability in areas such as tunnels, urban canyons, and indoors a new approach using least squares support vector machine (LS-SVM) andH∞filter (HF) for integration of INS/WSN is proposed. In the integrated system, HF estimates the errors of position and velocity while the signals in WSNs are available. Meanwhile, the compensation model is trained by LS-SVM with corresponding HF states. Once outages of the signals in WSNs, the model is used to correct INS solution as HF does. Moreover, due to device reasons, there are slight fluctuations in sampling period in practice. For overcoming this problem of integrated navigation, the theoretical analysis and implementation of HF for an integrated navigation system with stochastic uncertainty are also given. Simulation shows the performance of HF is more robust compared with INS-only solution and Kalman filter (KF) solution, and the prediction of LS-SVM has the smallest error compared with INS-only and back propagation (BP), the improvement is particularly obvious.

Sensor Dynamic Modeling Based on LS-SVM and NGA

2008 ◽  
Vol 381-382 ◽  
pp. 439-442
Author(s):  
Qi Wang ◽  
Zhi Gang Feng ◽  
K. Shida
Keyword(s):  
Genetic Algorithm ◽  
Neural Networks ◽  
Support Vector Machine ◽  
Least Squares ◽  
Dynamic Model ◽  
Dynamic Modeling ◽  
Support Vector ◽  
Local Minima ◽  
Highly Nonlinear ◽  
Sharing Function

Least squares support vector machine (LS-SVM) combined with niche genetic algorithm (NGA) are proposed for nonlinear sensor dynamic modeling. Compared with neural networks, the LS-SVM can overcome the shortcomings of local minima and over fitting, and has higher generalization performance. The sharing function based niche genetic algorithm is used to select the LS-SVM parameters automatically. The effectiveness and reliability of this method are demonstrated in two examples. The results show that this approach can escape from the blindness of man-made choice of LS-SVM parameters. It is still effective even if the sensor dynamic model is highly nonlinear.

MULTI-LEVEL DOCUMENT IMAGE SEGMENTATION USING MULTI-LAYER PERCEPTRON AND SUPPORT VECTOR MACHINE

2012 ◽  
Vol 26 (06) ◽  
pp. 1253002 ◽  
Author(s):  
YAN ZHANG ◽  
BIN YU ◽  
HAI-MING GU
Keyword(s):  
Support Vector Machine ◽  
Image Segmentation ◽  
Research Area ◽  
Document Image ◽  
Support Vector ◽  
Multi Layer Perceptron ◽  
Document Image Analysis ◽  
Important Research Area ◽  
Multi Level ◽  
Document Image Segmentation

Document image segmentation is an important research area of document image analysis which classifies the contents of a document image into a set of text and non-text classes. Previous existing methods are often designed to classify text and halftone therefore they perform poorly in classifying graphics, tables and circuit, etc. In this paper, we present a robust multi-level classification method using multi-layer perceptron (MLP) and support vector machine (SVM) to segment the texts from non-texts and thereafter classify them as tables, graphics and halftones. This method outperforms previously existing methods by overcoming various issues associated with the complexity of document images. Experimental results prove the effectiveness of our proposed method. By virtue of our multi-level classification approach, the text components, halftone components, graphic components and table components are accurately classified respectively which would highly improve OCR accuracy to reduce garbage symbols as well as increase compression ratio thereafter simultaneously.

scholarly journals Recognition of Human Emotion Detection and Annotation, using Local Descriptor and Support Vector Machine

2019 ◽  
Vol 9 (2) ◽  
pp. 2838-2843
Keyword(s):  
Support Vector Machine ◽  
Object Detection ◽  
Dynamic Environment ◽  
Detection Algorithm ◽  
Research Area ◽  
Multimedia Data ◽  
Video Annotation ◽  
Support Vector ◽  
People Detection ◽  
Local Descriptor

In multimedia data analysis, video tagging is the most challenging and active research area. In which finding or detecting the object with the dynamic environment is most challenging. Object detection and its validation are an essential functional step in video annotation. Considering the above challenges, the proposed system designed to presents the people detection module from a complex background. Detected persons are validated for further annotation process. Using publically available dataset for module design, Viola-Jones object detection algorithm is used for person detection. Support Vector Machine (SVM) authenticate the detected object/person based on it local features using Local Binary Pattern (LBP). The performance of the proposed system presents given architecture is effectively annotating the detected people emotion.

A rapid seismic data calibration technique using integrated micro-electro-mechanical system inertial sensor groups for a 3C vertical seismic profile

Geophysics ◽  
2017 ◽  
Vol 82 (6) ◽  
pp. P109-P118
Author(s):  
Huailiang Li ◽  
Xianguo Tuo ◽  
Tong Shen ◽  
Mark Julian Henderson ◽  
Jérémie Courtois
Keyword(s):  
Mechanical System ◽  
Seismic Data ◽  
Inertial Sensors ◽  
Inertial Sensor ◽  
Low Cost ◽  
Measuring Method ◽  
Seismic Profile ◽  
Micro Electro Mechanical System ◽  
Vertical Seismic Profile ◽  
Sensor Package

Calibration of 3C vertical seismic profile (VSP) data is an exciting challenge because the orientation of the tool is random when only seismic data are considered. We have augmented the sensor package on the VSP tool with micro-electro-mechanical system (MEMS) inertial sensors and applied a gesture measuring method to determine the tool orientation and calibration. This technique can quickly produce high precision, orientation, and angle information when integrated with the seismometer. The augmented sensor package consists of a low-cost triaxial MEMS gyroscope, an electronic compass, and an accelerometer. The technique to process the gesture information is based on the OpenGL software for 3D modeling. We have tested this approach on a large number of field data sets and it appeared to be faster and more reliable than other approaches.

scholarly journals Performance Evaluation of Support Vector Machine Algorithm for Human Gesture Recognition

2020 ◽  
pp. 204-210
Author(s):  
Vina Ayumi
Keyword(s):  
Support Vector Machine ◽  
Performance Evaluation ◽  
Gesture Recognition ◽  
Research Area ◽  
Human Motion ◽  
Support Vector ◽  
Body Movements ◽  
Average Accuracy ◽  
Svm Algorithm ◽  
Human Motion Recognition

Research on human motion gesture recognition has been widely used for several technological devices to support monitoring of human-computer interaction, elderly people and so forth. This research area can be observed by conducting experiments for several body movements, such as hand movements, or body movements as a whole. Many methods have been used for human motion gesture recognition in previous studies. This paper attempted to collect data of performance evaluation of support vector machine algorithms for human motion recognition. We developed research methodology that is adapted PRISMA. This methodology is consisted of four main steps for reviewing scientific articles, including identification, screening, eligibility and inclusion criteria. After we obtained result of systematic literature review. We also conducted pilot study of SVM implementation for human gesture recognition. Based on the previous study result, the accuracy performance of vector machine algorithms for body gesture dataset is between 82.88% - 99.92% and hand gesture dataset 88.24% - 95.42%. Based on our pilot experiment, recognition accuracy with the SVM algorithm for human gesture recognition achieved 94,50% (average) accuracy.

Sign in / Sign up

Export Citation Format

Share Document