International Journal of Modeling, Simulation, and Scientific Computing

Cyber-physical systems (CPS) tightly integrate cyber and physical components and transcend discrete and continuous domains. It is greatly desired that the synergy between cyber and physical components of CPS is explored even before the complete system is put together. Virtualization has potential to play a significant role in exploring such synergy. In this paper, we propose a CPS virtualization approach based on the integration of virtual machine and physical component emulator. It enables real software, virtual hardware, and virtual physical components to execute in a holistic virtual execution environment. We have implemented this approach using QEMU as the virtual machine and Matlab/Simulink as the physical component emulator, respectively. To achieve high-fidelity between the real system and its virtualization, we have developed a strategy for synchronizing the virtual machine and the physical component emulator. To evaluate our approach, we have successfully applied it to real-world control systems. Experiments results have shown that our approach achieves high-fidelity in capturing dynamic behaviors of the entire system. This approach is promising in enabling early development of cyber components of CPS and early exploration of the synergy of cyber and physical components.

Astrocyte plays a pivotal role in synaptic transmission with neuron. It maintains the ionic concentration in synapse to regulate signals from one cell to another. Calcium known as second messenger plays an important role in signal transduction. There are so many physiological processes that affect the cytosolic calcium concentration [Ca²⁺]_i like calcium buffering, flow of calcium ion through channels, etc. The modeling of calcium signaling in astrocytes has become more sophisticated. The modeling effort has provided insight to understand the cell contraction. Main objective of this work is to study the effect of voltage-gated calcium channel on calcium profile under excess buffer approximation in the form of diffusion equation. A mathematical model is developed in the form of diffusion equation for the calcium profile. The model incorporates the important physiological parameter like diffusion coefficient association rate constant, etc. Appropriate boundary conditions have been framed. Finite element method is employed to solve the problem. A MATLAB program has been developed for the entire problem and simulated to compute the numerical results.

This paper presents structures and principles of a fast stockpile simulation engine that enhances automatic material process by providing accurate digital information of material quality distribution within stockpiles. Compared with traditional measurement-based approaches, our simulation technology reduces operation cost by eliminating frequent physical measuring; improves operation accuracy through all-time whole-area contour simulation (instead of measuring a few positions of material layers now and then); and speeds the process through fast simulation and instant response to quality information query. Based on simplified grain dynamics and cellular automata models, the technology can achieve real-time/super-real-time simulation, which is critical to be adopted by the industry. Application of the technology in analyzing mineral ores handling is also discussed.

In this paper the CNC machining of St52 was modeled using an artificial neural network (ANN) in the form of a four-layer multi-layer perceptron (MLP). The cutting parameters used in the model were cutting fluid flow, feed rate, spindle speed and the depth of cut and the model output was the tool life. For obtaining more accuracy and spending less time Taguchi design of experiment (DOE) has been used and correlation between the output of the ANN and the experimental results was 96%. Further optimization process has been done by use of a genetic algorithm (GA). After optimization process tool life was increased about 8% equal to 33 min and was corroborated by experimental tests. This demonstrates that the coupling of an ANN with the GA optimization technique is a valid and useful approach to use.

Numerical study of multiphase fluid flows require mathematical methods for distinguishing interface between two fluids. The volume of fluid (VOF) method is one of such method which takes care of fluid shape in a local domain and reconstructs the interface from volume fraction of one fluid. Maintaining sharp interface during reconstruction is a challenging task and geometrical approach of VOF method better suits for incompressible fluids. This paper provides a complete mathematical discussion of extended form of VOF method using a approach known as piecewise linear interface calculation (PLIC). An analytical relation between volume fraction and interface position has been explored with the help of primitive geometrical shapes. The method with this analytical relation has been applied to multiphase fluid flow benchmark problems and found to be in good agreement.

This paper presents a high-throughput memory efficient decoder for low density parity check (LDPC) codes in the high-rate wireless personal area network application. The novel techniques which can apply to our selected LDPC code is proposed, including parallel blocked layered decoding architecture and simplification of the WiGig networks. State-of-the-art flexible LDPC decoders cannot simultaneously achieve the high throughput mandated by these standards and the low power needed for mobile applications. This work develops a flexible, fully pipelined architecture for the IEEE 802.11ad standard capable of achieving both goals. We use Real Time–Performance Evaluation Process Algebra (RT-PEPA) to evaluate a typical LDPC Decoder system's performance. The approach is more convenient, flexible, and lower cost than the former simulation method which needs to develop special hardware and software tools. Moreover, we can easily analyze how changes in performance depend on changes in a particular mode by supplying ranges for parameter values.

The variational grid generation method is a powerful tool for generating structured convex grids on irregular simply connected domains whose boundary is a polygonal Jordan curve. Several examples that show the accuracy of a finite difference approximation to the solution of a Poisson equation using this kind of structured grids have been recently reported. In this paper, we compare the accuracy of the numerical solution calculated using those structured grids and finite differences against the solution obtained with Delaunay-like triangulations on irregular regions.

Large-scale artificial societies with millions or billions of agents call for high-performance parallel simulation. Prevailing supercomputers with thousands of CPUs and GPUs make it possible to carry out such simulation. The key is to distribute large-scale agents to massive cores of CPUs and GPUs properly for parallel computing with efficient communication and synchronization. For simplicity and efficiency, a modified discrete event system specification (DEVS) is proposed for large-scale artificial society modeling and parallelism is exploited in agent models because similar agents usually share similar behaviors. Through phased synchronization, a two-tier parallel simulation engine is designed with support of MPI and OpenCL where GPU is used as coprocessor. One-sided communication is used for reflection of remote simulation objects and message passing between processes. A general kernel function prototype is elaborately designed and conditionally compiled for execution on both CPU and GPU. An artificial society for epidemic study is used to test the performance on a supercomputer with 1024 CPU cores and 1792 GPU cores. The speedup reaches 3512 for even 2 billion agents with GPU acceleration which is far over 701 when only CPUs are used. It turns out feasible for parallel simulation of large-scale artificial society with GPU as coprocessor.

An implementation of the standard collocation method based on polynomial interpolation is presented in a matrix framework in this paper. The underlying differentiation matrix can be partitioned to yield a superconvergent implicit multistep-like method to solve the initial value problem numerically. The first- and second-order versions of this method are L-stable.