scholarly journals Identification of Ship Dynamics Model Based on Sparse Gaussian Process Regression with Similarity

Symmetry ◽  
2021 ◽  
Vol 13 (10) ◽  
pp. 1956
Author(s):  
Gang Chen ◽  
Wei Wang ◽  
Yifan Xue
Keyword(s):  
Gaussian Process ◽  
Kernel Method ◽  
Fluid Dynamic ◽  
Gaussian Process Regression ◽  
Nonparametric Model ◽  
Dynamics Model ◽  
Large Sample ◽  
Model Based ◽  
Ship Dynamics ◽  
Parameterized Model

The system identification of a ship dynamics model is crucial for the intelligent navigation and design of the ship’s controller. The fluid dynamic effect and the complicated geometry of the hull surface cause a nonlinear or asymmetrical behavior, and it is extremely difficult to establish a ship dynamics model. A nonparametric model based on sparse Gaussian process regression with similarity was proposed for the dynamic modeling of a ship. It solves the problem, wherein the kernel method is difficult to apply to big data, using similarity to sparse large sample datasets. In addition, the experimental data of the KVLCC2 ship are used to verify the validity of the proposed method. The results show that sparse Gaussian process regression with similarity can be applied to the learning of a large sample data, in order to obtain ship motion prediction with higher accuracy than the parameterized model. Moreover, in the case of sensor signal loss, the identified model continues to provide accurate ship speed and trajectory information in the future, and the maximum prediction error of the motion trajectory within 100 s is only 0.59 m.

scholarly journals An Efficient Radio Map Updating Algorithm based on K-Means and Gaussian Process Regression

2018 ◽  
Vol 71 (5) ◽  
pp. 1055-1068 ◽  
Author(s):  
Jianli Zhao ◽  
Xiang Gao ◽  
Xin Wang ◽  
Chunxiu Li ◽  
Min Song ◽  
...  
Keyword(s):  
Computational Complexity ◽  
Gaussian Process ◽  
Predictive Model ◽  
Gaussian Process Regression ◽  
Environment Changes ◽  
Model Based ◽  
Map Updating ◽  
Mean Function ◽  
Radio Map ◽  
Indoor Localisation

Fingerprint-based indoor localisation suffers from influences such as fingerprint pre-collection, environment changes and expending a lot of manpower and time to update the radio map. To solve the problem, we propose an efficient radio map updating algorithm based on K-Means and Gaussian Process Regression (KMGPR). The algorithm builds a Gaussian Process Regression (GPR) predictive model based on a Gaussian mean function and realises the update of the radio map using K-Means. We have conducted experiments to evaluate the performance of the proposed algorithm and results show that GPR using the Gaussian mean function improves localisation accuracy by about 13·76% compared with other functions and KMGPR can reduce the computational complexity by about 7% to 20% with no obvious effects on accuracy.

A data driven BRDF model based on Gaussian process regression

2013 ◽  
Author(s):  
Zhuang Tian ◽  
Dongdong Weng ◽  
Jianying Hao ◽  
Yupeng Zhang ◽  
Dandan Meng
Keyword(s):  
Gaussian Process ◽  
Gaussian Process Regression ◽  
Data Driven ◽  
Model Based

scholarly journals Minimum Mode Saddle Point Searches Using Gaussian Process Regression with Inverse-Distance Covariance Function

2019 ◽  
Author(s):  
Olli-Pekka Koistinen ◽  
Vilhjálmur Ásgeirsson ◽  
Aki Vehtari ◽  
Hannes Jónsson
Keyword(s):  
Saddle Point ◽  
Gaussian Process ◽  
Process Model ◽  
Energy Surface ◽  
Saddle Points ◽  
Gaussian Process Regression ◽  
Dissociative Adsorption ◽  
Initial Guess ◽  
Initial State ◽  
Model Based

The minimum mode following method can be used to find saddle points on an energy surface by following a direction guided by the lowest curvature mode. Such calculations are often started close to a minimum on the energy surface to find out which transitions can occur from an initial state of the system, but it is also common to start from the vicinity of a first order saddle point making use of an initial guess based on intuition or more approximate calculations. In systems where accurate evaluations of the energy and its gradient are computationally intensive, it is important to exploit the information of the previous evaluations to enhance the performance. Here, we show that the number of evaluations required for convergence to the saddle point can be significantly reduced by making use of an approximate energy surface obtained by a Gaussian process model based on inverse inter-atomic distances, evaluating accurate energy and gradient at the saddle point of the approximate surface and then correcting the model based on the new information. The performance of the method is tested with start points chosen randomly in the vicinity of saddle points for dissociative adsorption of an H2 molecule on the Cu(110) Surface and three gas phase chemical reactions.<br>

scholarly journals Localization Reliability Improvement Using Deep Gaussian Process Regression Model

Sensors ◽  
2018 ◽  
Vol 18 (12) ◽  
pp. 4164 ◽  
Author(s):  
Fei Teng ◽  
Wenyuan Tao ◽  
Chung-Ming Own
Keyword(s):  
Gaussian Process ◽  
Dynamic Environment ◽  
Gaussian Process Regression ◽  
Computation Time ◽  
Indoor Positioning ◽  
Location Estimation ◽  
Reference Points ◽  
Nonparametric Model ◽  
Positioning System ◽  
Positioning Accuracy

With the widespread use of the Global Positioning System, indoor positioning technology has attracted increasing attention. Many systems with distinct deployment costs and positioning accuracies have been developed over the past decade for indoor positioning. The method that is based on received signal strength (RSS) is the most widely used. However, manually measuring RSS signal values to build a fingerprint database is costly and time-consuming, and it is impractical in a dynamic environment with a large positioning area. In this study, we propose an indoor positioning system that is based on the deep Gaussian process regression (DGPR) model. This model is a nonparametric model and it only needs to measure part of the reference points, thus reducing the time and cost required for data collection. The model converts the RSS values into four types of characterizing values as input data and then predicts the position coordinates using DGPR. Finally, after reinforcement learning, the position coordinates are optimized. The authors conducted several experiments on a simulated environment by MATLAB and physical environments at Tianjin University. The experiments examined different environments, different kernels, and positioning accuracy. The results showed that the proposed method could not only retain the positioning accuracy, but also save the computation time that is required for location estimation.

A genetic Gaussian process regression model based on memetic algorithm

2013 ◽  
Vol 20 (11) ◽  
pp. 3085-3093 ◽  
Author(s):  
Le Zhang ◽  
Zhong Liu ◽  
Jian-qiang Zhang ◽  
Xiong-wei Ren
Keyword(s):  
Regression Model ◽  
Gaussian Process ◽  
Memetic Algorithm ◽  
Gaussian Process Regression ◽  
Model Based

scholarly journals Pointing error compensation of electro-optical detection systems using Gaussian process regression

2021 ◽  
Vol 12 ◽  
pp. 22
Author(s):  
Qijian Tang ◽  
Qingping Yang ◽  
Xiangjun Wang ◽  
Alistair B. Forbes
Keyword(s):  
Gaussian Process ◽  
Error Compensation ◽  
Optical Detection ◽  
Gaussian Process Regression ◽  
Parametric Models ◽  
Nonparametric Model ◽  
Important Indicator ◽  
Pointing Error ◽  
Detection Systems ◽  
Pointing Accuracy

Pointing accuracy is an important indicator for electro-optical detection systems, as it significantly affects the system performance. However, as a result of misalignment, nonperpendicularity in the manufacturing and assembly processes, as well as the sensor errors such as camera distortion and angular sensor error, the pointing accuracy is significantly affected. These errors should be compensated before using the system. Parametric models are firstly proposed to compensate for the errors, whilst the semi-parametric models with the nonlinearity added are also put forward. Both methods should analyse the parametric part first, which is a complicated and inaccurate process. This paper presents a nonparametric model, without any prior information about mechanical dimensions, etc. It depends only on the test data. Gaussian Process regression is used to represent the relationship between data and predict the compensated output. The test results have shown that the regression variances have decreased by more than an order of magnitude, and the means have also been significantly reduced, with the pointing error well improved. The nonparametric model based on Gaussian Process is thus demonstrated to be an effective and powerful tool for the pointing error compensation.

Sign in / Sign up

Export Citation Format

Share Document