Kourosh Rahnamai

Model Predictive Neural Control of TCP Flow in AQM Network

Many research papers have been published on RED (random early detection) and variants of RED. Recently many articles have been presented on modeling a transmission control protocol (TCP) flow in an active queue management (AQM) of a bottlenecked network link (Dong Lin and R. Morris, 1997), (V. Misra, et al., 2000), (C. Hollot, et al., 2001). Classical control theories have also been applied to achieve or improve stability of the network flow (C. Hollot, et al., 2001). In this paper we present a neural network (NN) model predictive control (MPC) of TCP flows. We show the robust adaptive behavior of the MPC optimal controller under modeling errors and system dynamic changes. We also show the superior transient and steady state behavior as well as general stability of MPC as compared to the classical PI controller

Fuzzy Automatic Guitar Tuner

An automatic guitar tuner was successfully designed and developed using a fuzzy logic controller. The guitar tuner was implemented using Simulink and XPC real time kernel. The system acquires the signal from an electrical guitar and inputs the signal into a target PC running XPC. Using fast Fourier transforms (FFT), the system calculates the fundamental and harmonics of the played notes and compares it with the desired pattern. The frequency difference is used as an input to a fuzzy logic controller that automatically adjusts the tension of the desired string.

Fan and Plate Apparatus for Illustrating Stability Concepts Using XPC Rapid-Prototyping Environment

In this study, a fan and plate apparatus is used to demonstrate the aerodynamic resonance phenomena and stability concepts. A series of controllers such as P, PD and PID are used to illustrate the aerodynamic behavior of this system. This laboratory apparatus can be used in an undergraduate or graduate controls course to illustrate and reinforce stability and controller design concepts. The xPC Rapid Prototyping Toolbox of MATLAB® is used for a fast and flexible method to implement a variety of controllers using a high-level programming language such as Simulink® .Copyright

Quadrotor drones thrust measurement apparatus

The UAV field has been growing rapidly and is poised for even larger growth in the coming years. While the military has been using UAV aircrafts for years, there is potentially room for growth in the civilian market. UAVS with vertical takeoff and landing (VTOL) capabilities have generated much interest. For these reasons, a few years ago, Quadrotor UAV undergraduate research and development studies were started at Western New England University Controls Laboratory. One major design concern of any Quadrotor drone development platform is how to measure the torque developed at the output shaft of each of the Quadrotor motors. Fraction horsepower (HP) DC motor dynamometers are hard to find and very expensive. Research publications on how to measure and estimate a brushed or brushless DC motor characteristic parameter methodology are also too complex for undergraduate students to understand in order for them to design and develop a measurement apparatus [1-6]. In this paper we will discuss the design and development of an affordable torque measurement apparatus to measure efficiency and estimate characteristic parameters of fraction horsepower brushed and brushless DC motors.

High reliability neural networks structure with application to spacecraft ASMS tone detection

The authors show that the research on N-version high-reliability software structures can be extended to neural network architectures. In addition, we explore the possibility of applying this structure to a spacecraft tracking problem. One such system is the Automated Spacecraft Monitoring System (ASMS), a beacon-monitoring or detection system. Four neural networks, each trained for various operating environments, are implemented in an N-version structure. The results of the networks are combined to form a composite outcome. The combined outcome is used as part of a hypothesis testing procedure to distinguish between the presence or absence of the beacon signal. The results show that any of a number of composite outcomes outperforms the use of any single neural network. Further, the simple average of network results provides the composite outcome with best performance.

Fuzzy supervised roll axis controller for a hovering system

Vertical take off and landing aircrafts have been the subject of research for many years [1–5]. The Boeing CV-22 Osprey is one such aircraft. Takeoff and landing phases of these aircrafts create very challenging control problems. Having two massive engines at the wingtips of the airplane creates an inherently unstable system. Added to this challenge is an inherent delay in pilot demand for engine thrust changes. In this paper we will present an intelligent fuzzy supervisory agent [6,8] that monitors and modifies the gain scheduled controllers at high roll angles. This fuzzy agent modifies the thrust level and the controller parameters. To model the roll axis behavior of such aerial vehicles two electric motors with propellers are mounted at the end of a beam with an almost frictionless pivoting point at the center. A classical Gain scheduled PID controller is used as the primary controller for this system.

Fuzzy logic regulated power supply for a UHF digital television transmission

In this work, we show that the fuzzy logic controller can be used to significantly improve the performance of a power supply regulator over conventional control techniques. The existing linear regulator circuit used in the current design has excessive overshoot and ringing in response to step changes, particularly during initial turn-on into a cold (very low impedance) filament. The linear regulator compensates for gradual changes in the load, but an external system must apply the reference voltage gradually to prevent excessive overshoot. For these reasons, a fuzzy controller is an excellent choice to provide the needed nonlinear regulation. This regulator is used in high power amplifiers produced by Thales Broadcast and Multimedia, a manufacturer of UHF television transmitters located in Southwick, Massachusetts.

State space design cycle and hardware in the loop testing and verification

Advances in real-time hardware in the loop (HIL) boards and software has provided an excellent opportunity to implement and test control algorithms rapidly on real systems [1-7]. Hardware in the loop development allows for rapid development, testing and verification of control systems. This paper, using a servo system as an example, clearly explores modern control systems design cycle, which includes real-time verification of the controller. This methodology illustrates design cycle from system specification, system constraints to controller design, HIL testing and verification of the developed controller. This rapid prototyping strategy to develop and test controllers can be used in industry as well as academia. In an academic environment, it helps student relate, better understand and implement theoretical concepts on real systems. Experiments described in this paper use a realtime hardware in the loop platform, which consists of a DC-servo control system, MATLAB, Simulink, dSPACE real-time hardware and ControlDesk software. This paper shows how this type of platform can is be used to verify system models and test designed state space controllers.

Systems engineering in industry internship and academic projects

Large complex Industrial projects most often are interdisciplinary and require a complex systems engineering approach to guarantee success. In this paper we present a successful implementation of a true systems engineering approach to a summer internship program in industry. We exported the methods and lessons learned to an academic environment. For each project a group of six to fifteen students were selected to solve a practical industry problem and produce a detailed design for a specified project. Different aspects of the project plan were assigned to members of each group who were the most qualified or who expressed interest in a specific area of specialization. Three industrial and one academic implementation of this method are explained in this paper.

Thermal Engine with Metastable Power Extraction Step (TEMPES)

In this paper we present the design and development process of a high efficiency Thermal Engine with Metastable Power Extraction Step (TEMPES) for high efficiency, low temperature heat recovery. TEMPES is a closed cylinder 2-cycle engine which is designed to illustrate energy recovery from low grade industrial wasted heat. The basic operating principle of the TEMPES engine is a closed cylinder piston assembly that converts heat energy in a working fluid to mechanical energy using innovative chemical processes to extract greater than 50% more energy from a thermal energy source relative to traditional heat engines restricted to bulk thermodynamic equilibrium processes.