Minh Pham

A Model for the Forward Problem in Magnetic Induction Tomography Using Boundary Integral Equations

We propose a new formulation for the forward problem in magnetic induction tomography (MIT). We formulate the problem in terms of interior and exterior boundary integral equations (BIEs), subject to appropriate boundary conditions. We then transform a standard exterior BIE involving the magnetic vector potential to a BIE involving the electric fields. This transformation eliminates two boundary conditions involving the magnetic vector potential and its normal derivative. This greatly reduces the computational complexity of the model. Here, we compare numerical solutions of the model to analytical solutions.

Imaging the solidification of molten metal by eddy currents: I

This paper presents an eddy-current-based technique for imaging the solidification of molten metal inside a metal pipe. This technique is motivated by the fact that the ac (alternating current) impedance seen from a solenoid placed outside the pipe is dependent on the distribution of the solidification inside the pipe provided that the ac frequency is properly chosen and the overall thickness of the pipe and solidification is small. This paper first establishes a simplified mathematical model based on the fundamental electromagnetic theory, which reveals the exact relationship between the solidification inside the pipe and the scattered field outside the pipe. Based on this model, an iterative algorithm for reconstructing the solidification distribution is then developed. (Some figures in this article appear in colour in the electronic version; see www.iop.org)

Real-Time Optimal Control of River Basin Networks

River basins are key components of water supply grids. River basin operators must handle a complex set of objectives including runoff storage, flood control, supply for consumptive use, hydroelectric power generation, silting management, and maintenance of river basin ecology. At present, operators rely on a combination of simulation and optimization tools to help them make operational decisions. The complexity associated with this approach makes it suitable for long term planning but not daily or hourly operation. The consequence is that between longerterm optimized operation points, river basins are largely operated in open loop. This leads to operational inefficiencies most notably wasted water and poor ecological outcomes. This paper proposes a systematic approach using optimal control based on simple low order models for the real-time operation of entire river basin networks.

Routing and localization for extended lifetime in data collection wireless sensor networks

Information collected by wireless sensor networks could be essential to community network services. We address two fundamental issues of wireless sensor networks - routing and localization in data collection applications. In particular, by utilizing special characteristics of the applications, energy-efficient methods for routing and localization are proposed. Without any path search, the proposed routing method is able to find the optimal paths which reduce radio activities and balance energy consumption among sensor nodes. The proposed localization method produces the relative locations of nodes without the need for any GPS-capable nodes. The proposed routing and localization methods are, therefore, effective in extending the network lifetime.

Computation of 3-D Sensitivity Coefficients in Magnetic Induction Tomography Using Boundary Integral Equations and Radial Basis Functions

This paper presents a method for the numerical computation of 3-D sensitivity coefficients of a target object in magnetic induction tomography (MIT). The sensitivity coefficient at a point is defined as the dot product of electromagnetic fields produced by unit current flowing in the excitation and the detector coil. In this paper, the fields are governed by a set of boundary integral equations (BIEs). Numerical results demonstrate that the fields on the boundary and interior volume domain of the target can be accurately represented by radial basis functions (RBFs). The paper compares numerical solutions of the BIEs based on RBFs with analytical solutions and boundary element solutions.

Model predictive control of Murray-darling basin networks

River basins are the most significant component in water supply grids and are under increasing pressure from competing demands for fresh water. However, unlike energy grids which are managed very efficiently using closed-loop operation, water grids, and river basins in particular, are largely open-loop systems. One reason is the difficulty associated with developing suitable models and feedback controllers. This paper proposes a systematic approach using model predictive control based on simple low order models for the real-time operation of entire river basin networks.

Robust Multipath Links for Wireless Sensor Networks in Irrigation Applications

Wireless sensor networks for irrigation applications normally consist of low-cost and low-power nodes deployed in a harsh environment. It turns out that the radio links between the coordinating node and actuating nodes are more critical and are expected to be more robust to radio interference and node failure than links between the coordinating node and sensing nodes. We present an efficient method to create robust links for irrigation sensor networks built on the ZigBee specification. In particular, multipaths between the coordinating node and actuating nodes are created and used to enhance the critical links. It is shown that, by utilizing the properties of addresses of ZigBee networks, those multipaths can be dynamically created and released adapting to topology changes without any path search activities. Simulation results confirm the robustness of the proposed multipath links.

Real-time optimization of irrigation scheduling in agriculture

A simulation optimization approach is developed for managing irrigation operations in agriculture. The approach combines a highly complex crop growth and yield simulation model with an optimization algorithm to calculate an irrigation sequence that maximizes crop yield. The approach overcomes the limitations of many existing irrigation scheduling optimization methods that rely on simpler but more inaccurate crop growth and yield models to ensure computational tractability. The performance of the approach is demonstrated through simulation.

On orientations of directional antennas in wireless sensor network

The paper presents a model for the total Shannon capacity of a network as a function of antenna orientations. The resulting function is neither concave nor convex and is multi-extrema with multiple local optima. An optimum solution for the orientations of the directional antennas leads to the global maximization of the total channel capacity. This global optimization problem cannot be solved by classical nonlinear programming technique. The objective function is shown to be Lipschitzian and is optimized by a Lipschitz optimization algorithm. The algorithm is a realization of the generic branch and bound methodology. Simulation results are reported and discussed.

Comparison of modelling approaches for the eddy current problem as applied to the geometry of a taphole

An eddy current tomography system is under development to monitor the condition of metallurgical tapholes. In this paper, the results of two approaches for solving the forward eddy current problem are presented. The results of these simulations are compared against corresponding experimental measurements taken using a custom-built measurement system and a commercial LCR meter. These comparisons showed that there was good agreement between the measured and predicted responses for both the phase angle and magnitude components of the response. Additional modelling conducted showed that the measurement system will be sensitive to the changes that will occur as the wear profile of the taphole refractory changes over time.