ADMM-based adaptive sampling strategy for nonholonomic mobile robotic sensor networks
- Authors: Le, Viet-Anh , Nguyen, Linh , Nghiem, Truong
- Date: 2021
- Type: Text , Journal article
- Relation: IEEE Sensors Journal Vol. 21, no. 13 (2021), p. 15369-15378
- Full Text:
- Reviewed:
- Description: This paper discusses the adaptive sampling problem in a nonholonomic mobile robotic sensor network for efficiently monitoring a spatial field. It is proposed to employ Gaussian process to model a spatial phenomenon and predict it at unmeasured positions, which enables the sampling optimization problem to be formulated by the use of the log determinant of a predicted covariance matrix at next sampling locations. The control, movement and nonholonomic dynamics constraints of the mobile sensors are also considered in the adaptive sampling optimization problem. In order to tackle the nonlinearity and nonconvexity of the objective function in the optimization problem we first exploit the linearized alternating direction method of multipliers (L-ADMM) method that can effectively simplify the objective function, though it is computationally expensive since a nonconvex problem needs to be solved exactly in each iteration. We then propose a novel approach called the successive convexified ADMM (SC-ADMM) that sequentially convexify the nonlinear dynamic constraints so that the original optimization problem can be split into convex subproblems. It is noted that both the L-ADMM algorithm and our SC-ADMM approach can solve the sampling optimization problem in either a centralized or a distributed manner. We validated the proposed approaches in 1000 experiments in a synthetic environment with a real-world dataset, where the obtained results suggest that both the L-ADMM and SC-ADMM techniques can provide good accuracy for the monitoring purpose. However, our proposed SC-ADMM approach computationally outperforms the L-ADMM counterpart, demonstrating its better practicality. © 2001-2012 IEEE.
Efficient evaluation of remaining wall thickness in corroded water pipes using pulsed Eddy current data
- Authors: Nguyen, Linh , Miro, Jaime Valls
- Date: 2020
- Type: Text , Journal article
- Relation: IEEE Sensors Journal Vol. 20, no. 23 (2020), p. 14465-14473
- Full Text:
- Reviewed:
- Description: In order to analyse failures of an ageing water pipe, some methods such as the loss-of-section require remaining wall thickness (RWT) along the pipe to be fully known, which can be measured by the magnetism based non-destructive evaluation sensors though they are practically slow due to the magnetic penetrating process. That is, fully measuring RWT at every location in a water pipe is not really practical if RWT inspection causes disruption of water supply to customers. Thus, this paper proposes a new data prediction approach that can increase amount of RWT data of a corroded water pipe collected in a given period of time by only measuring RWT on a part (e.g. 20%) of the total pipe surface area and then employing the measurements to predict RWT at unmeasured area. It is proposed to utilize a marginal distribution to convert the non-Gaussian RWT measurements to the standard normally distributed data, which can then be input into a 3-dimensional Gaussian process model for efficiently predicting RWT at unmeasured locations on the pipe. The proposed approach was implemented in two real-life in-service pipes, and the obtained results demonstrate its practicality. © 2001-2012 IEEE.
Least square and Gaussian process for image based microalgal density estimation
- Authors: Nguyen, Linh , Nguyen, Dung , Nghiem, Truong , Nguyen, Thang
- Date: 2022
- Type: Text , Journal article
- Relation: Computers and Electronics in Agriculture Vol. 193, no. (2022), p.
- Full Text: false
- Reviewed:
- Description: Efficiently monitoring microalgal density in real time is critical in closed systems of cultivating algae. In the monitoring methods proposed in the literature, image based techniques present practically potential since they are nondestructive and more biosecured. However, in the existing image analysis methods, parameters of the color-to-grayscale conversion formulae are predefined and only applicable to monitor some specific microalgae strains. Therefore, in this paper we propose a generic approach based on least square to optimize those parameters, which are data-driven and can be used to monitor any type of microalgae. More importantly, apart from the widely used linear regression paradigm, we propose a nonlinear regression model based on Gaussian process to better learn relationship between data representation of measured images and densities of microalgae. The nonlinear regression model is then utilized to efficiently estimate density of algal species. The proposed approach was evaluated in the real-world dataset of Chlorella vulgaris microalgae, where the obtained results as compared with those obtained by some existing techniques demonstrate its effectiveness. © 2022 Elsevier B.V.
Multivariate versus univariate sensor selection for spatial field estimation
- Authors: Nguyen, Linh , Thiyagarajan, Karthick , Ulapane, Nalika , Kodagoda, Sarath
- Date: 2021
- Type: Text , Conference paper
- Relation: 16th IEEE Conference on Industrial Electronics and Applications, ICIEA 2021 p. 1187-1192
- Full Text: false
- Reviewed:
- Description: The paper discusses the sensor selection problem in estimating spatial fields. It is demonstrated that selecting a subset of sensors depends on modelling spatial processes. It is first proposed to exploit Gaussian process (GP) to model a univariate spatial field and multivariate GP (MGP) to jointly represent multivariate spatial phenomena. A Matérn cross-covariance function is employed in the MGP model to guarantee its cross-covariance matrices to be positive semi-definite. We then consider two corresponding univariate and multivariate sensor selection problems in effectively monitoring multiple spatial random fields. The sensor selection approaches were implemented in the real-world experiments and their performances were compared. Difference of results obtained by the univariate and multivariate sensor selection techniques is insignificant; that is, either of the methods can be efficiently used in practice. © 2021 IEEE.