Projects

DMon Process Monitor

DMon is a small process daemonization and monitorization tool, àla daemontools: it will spawn long-running programs and (optionally) logging programs connected by their standard output/input streams. The nice thing about DMon is that it is very easy to use, because it does not need support files on ... [More] disk and the most common process tweaks can be done with command line flags. Finally, it has some unique features, like the abilities of not really daemonizing (for interactive use), writing PID-files, running one-shot processes and specifying a maximum processing time. [Less]

metaweb-nanny

The Nanny was developed at Metaweb Technologies to greatly increase the reliability and traceability of the various Freebase.com production services. Its feature set includes: Auto-daemonisation with optional automatic restart. One nanny process can support multiple distinct children. Rotated ... [More] logging of child's standard IO streams and events such as starts, stops and restarts. Intelligent stopping of children. SIGTERM at first, then will try a SIGKILL if that doesn't work. In-built HTTP server for logs and statistics with both humanly-readable and machine-readable JSON output. Timed or periodic operation support. Similar conceptually to cron, but randomises the start time to avoid machines all running the operation at exactly the same time. [Less]

iscada3d

The system use a CAN fieldbus as a distributed control network. One of the nodes is a control unit based on RTLinux. From the desktop PC, and via Internet, the process can be monitored and managed. The main features of the platform are: Control Processes under Real-Time. Distributed control ... [More] topology due to the use of CAN fliedbus. Multiplatform desktop interface. 3D representation capability under Blender. [Less]

android-opc

A toolkit to browse an OPC server from a mobile device running the Android operating system.

XPCA-Server

XPCA-Server is experimental server for XPCA

g-task

This framework enables small to mid-sized organizations to create robust meta-processes which interface with common GIS and Image Processing systems. Python is the tool used to bind this framework to external engines. Code for binding G-Task to the Idrisi GIS system is provided.

lpcvd

IntroductionThe objective of this project is to establish a systematic framework for the design of feedback control system for real-time control of thin film microstructural defects using integration of multiscale models and measurements. Key elements include: Method for construction of ... [More] stochastic dynamic models, which are suitable for controller design and real-time implementation. Practical-implementable, feedback control systems based on nonlinear and predictive control theory. Multi-scale modeling of thin film growth describing the relationship between controllable process variables and thin film porosity. Construction of stochastic dynamic models that describe the effect of controllable variables on the evolution of thin film porosity using data from multiscale process model simulations. Design of estimation and feedback control systems for real-time thin film porosity minimization using the stochastic dynamic models and process measurements. Detailed works are listed below: Process simulator for thin film deposition (LPCVDSIM). This is based on kinetic Monte Carlo algorithm developed in \cite{Gillespie1976, Fichthorn1991, Shitara1992, Reese2001}. The simulator comes in two way (1) a set of classes that you can use in your code (2) a stand-alone binary file that your can integrated with Matlab or other common control developing platform. Matlab and general interface for this software is also provided with the simulator. Besides, GPU implementation is also considered. The interface of this simulator is designed so that it can be easily replaced with a real system. Algorithms planed to implement includes: solid-on-solid, 1D solid-on-solid, 2D triangular lattice, 1D real crystal structure, 2D Multiscale simulator combining 1D PDE gas phase with kMC simulator. Multiscale simulator combining 2D/3D CFD model with kMC simulator Control algorithms package (LPCVDCON). This will be implemented as a Matlab library and Simulink block library. The design principle of this library is that it requires few modification to apply in real system. Feedback controller using multiple small KMC simulator as in \cite{Lou2003a,Lou2003b} State feedback controller of surface roughness based on linear stochastic PDE as in \cite{Lou2005a, Lou2005b} State feedback controller of surface roughness based on nonlinear stochastic PDE as in \cite{Lou2006} Output feedback controller of surface roughness based on linear and nonlinear stochastic PDE as in \cite{Hu2008} Model predictive controller based on KSE equation for surface roughness, porosity and thickness Model predictive controller based on multiscale model for surface roughness, porosity and thickness Model predictive control of surface roughness variance System identification algorithms from KMC simulation results (LPCVDID). This part also come as a Matlab library. Algorithms planed to implemented includes: Deterministic linear model as in \cite{Siettos2003, Armaou2004, Varshney2006} Reduced-order approximation of the master equation as developed in \cite{Gallivan2004}. Empirical input-output models as used by \cite{Rusli2006} KSE model You can find more about this project from the proposal. MembersAdvisers: Panagotis D. Christofides, Orkoulas Markis Graduate students: Gangshi Hu, Xinyu Zhang, Jianqiao Huang For DeveloperWeekly Plan and Progress ReportsCode Structure and Interface SpecificationData Format [Less]