Research in Systems, Measurement, and Control at the School of Mechanical Engineering, Purdue University

Faculty members of the Systems, Measurement and Control Area in the School of Mechanical Engineering have diverse expertise in addressing the field’s major driving forces.   The systems that are a part of our everyday life in the 21st century will be an elegant mixture of the fruits of current research on integrating leading-edge measurements, actuation, electronics, control and computing capabilities, coupled with our increased understanding of mechanical system behavior.   Applications areas include transportation systems, manufacturing systems, biomedical devices, robotics, electro-hydraulic systems, and electro-mechanical systems.   This interdisciplinary group of engineers searches for solutions to these issues.

Engine Diagnostics

Advanced system identification and prognosis algorithms for diesel engines are being developed using information-theoretic metrics.  Algorithms are being applied to 16-cylinder diesel engines that are used in heavy-duty mining trucks. These algorithms will enable early fault detection leading to reduced down-time and lower maintenance costs.  For example, after assessing the  importance of each signal on engine health, the most informative sensor set can be chosen to generate cross-plots, identify data clusters, and quantify how these move with fault conditions.


Modeling/Control of a Robotic Endoscope via Undulated motion

An endoscope consists of fiber optics housed in flexible or semi-flexible tubing that can be inserted into a patient’s large intestine to provide an internal view of the patient without the need for a surgical opening.  Unfortunately, insertion and positioning of the endoscope is often difficult due to its flexible nature and the curved shape of the large intestine. 

Self-propelled robotic endoscopes have been proposed to alleviate the problem of insertion.  By emulating the undulating motion of a snake, the robotic endoscope can assist the surgeon via self-insertion and positioning.  While the dynamics of robotic snakes have been extensively covered in the literature, the majority of the designs presented thus far rely on the use of un-actuated wheels to provide a non-holonomic constraint necessary for forward motion.  Since wheeled motion inside the human body is impracticable, a new model and control methodology is necessary.

This research focuses on the development of a control-oriented model of the undulated motion of a snake robot moving through an enclosed environment of unknown stiffness.  Furthermore, a methodology for controlling the motion of the robot is sought in combination with the development of the model.  It is hoped that by removing the reliance of the un-actuated wheels to provide a non-holonomic constraint, a more appropriate model/control strategy for a self-propelled robotic endoscope can be developed.   

Modeling and Simulation of Navy Ship System of Systems

Manpower accounts for more than 80% of the operational costs of modern naval destroyers and aircraft carriers.  In order to accelerate transformation of ships for a variety of future missions, the U.S. Navy must develop information technologies for optimizing manpower (e.g., prognosis technologies, wireless PDAs and LANs, video surveillance sensor networks for threat detection).  The Navy currently uses Sea Trials to assess the impacts of new technologies on ships’ crews in an ad hoc manner at a cost of 100’s of millions of dollars per trial (e.g., ’06 Trident Sea Warrior).  This program aims to establish modeling and simulation resources to support technology decision making and optimize the planning of Sea Trials. 

The fundamental barrier to address in this work is the highly interconnected nature of crew, infrastructure, and business processes on a ship.  For example, consider the Arleigh Burke class DDG-51 shown below.  Because ships are highly interconnected, they exhibit emergent phenomena that may not be observed in Sea Trials.  We aim to model and simulate these phenomena to predict if communications, prognosis, and other technologies for infrastructure enable the ship’s crew to efficiently execute normal maintenance protocols and emergency procedures in the event of external threats as in the case of the U.S.S. Cole.  Other issues to address include the adaptive nature of ship environments and the uncertainties in models utilized to simulate the ship infrastructure and crew.


Arleigh Burke class DDG-51 guided missile destroyer; and example of interconnectivity among the various ship’s infrastructure.

Intelligent Control of Complex Systems

Many systems and manufacturing processes encountered in the real world are highly nonlinear and multivariable.  In such cases establishing accurate analytical models is often very difficult.   Despite such difficulties needs often arise to control such complex systems and processes.   Advances in artificial intelligence, soft computing and related scientific fields have brought new opportunities and challenges for researchers to deal with complex or/and uncertain problems and systems, which could not be solved by traditional methods.   Many existing approaches that have been developed for mathematically well-defined problems with accurate models may lack in autonomy and decision making ability and hence cannot provide adequate solutions under uncertain or dynamic environments. 


The SMAC Area faculty and students have developed intelligent control schemes (ICS) that provide a new approach to addressing those complex problems with uncertainties.   Intelligent control systems are defined with such attributes as: high degree of autonomy, reasoning under uncertainty, higher performance in a goal seeking manner, high level of abstraction, data fusion from a multitude of sensors, learning and adaptation in a heterogeneous environment, etc.  In this project, they developed a method that provides means of incorporating analytical models, empirical relationships and heuristic rules in a unified fashion, and searches for an optimal solution using a novel inferencing mechanism.  They also create models of the systems or processes to be controlled with generic modeling tools, such as radial basis function networks and fuzzy basis function networks based on the input-output data, based on a novel autonomous learning scheme.   These models are subsequently used to construct an intelligent control actions for control of various complex processes.   The intelligent control system utilizes a hierarchical control structure to provide the capabilities for controlling unknown or time-varying dynamic systems. 

Current Project Areas

Below is a list of current research projects in the Systems, Measurement and Control Areas:

Modeling and System Identification

  • Modeling and control of fuel cell power systems
  • Modeling of machining chatter
  • Autonomous modeling of complex manufacturing processes
  • Real-time predictive modeling and simulation for prognosis in smart ships
  • Real-time load and damage identification in filament wound rocket motor casings
  • Modeling and identification of seat head rest rattle in passenger bucket seat
  • Modeling and identification of morphing aircraft structure

Measurements and Diagnostics

  • Nonlinear observer design and neural networks for virtual sensing, modeling, fault detection, diagnostics, and adaptive robust fault-tolerant control.
  • Diesel engine diagnostics and prognostics using information-rich input signals
  • Estimating particulate load in a diesel particulate filter for regeneration control
  • Vehicle health management technologies
  • Integrated diagnostics and reliability forecasting for heterogeneous structures
  • A facility for theoretical and experimental environmental conditioning, modeling and prognostics of advanced heterogeneous structures
  • Sensing and diagnostics of electrical machines
  • Autonomous selection of sensors and sensor features for intelligent monitoring and diagnostics for manufacturing processes
  • Integrated prognostics health management technologies for commercial and defense systems
  • Integrated sensing and diagnostics for life cycle health management of gas turbine engines:  application to wire harnesses and connectors
  • Modeling and diagnostics of mechanically attached structural components
  • Diagnostics & prognostics for assessing vehicle products in real time with feedback for manufacturing to reduce conservatism

Control Theory

  • Nonlinear adaptive robust control theory
  • Multi-level fuzzy control
  • Multivariable intelligent control
  • Neural network-based adaptive control
  • Observer-based adaptive control

Control Applications

  • Engine controls
  • Energy-saving nonlinear control of electro-hydraulic systems
  • Intelligent and precision control of high-speed linear motor drive systems, machine tools, and piezo-electric actuators for precision manufacturing
  • Nonlinear control of high-density hard disk drives
  • Coordinated control of robot manipulators
  • Feedforward/feedback motion control for high-speed automation
  • High-speed motion control for flexible robotic manipulators
  • Precision control of piezoelectric actuators for scanning microscopes
  • Control of medical devices
  • Control of mechatronic devices