August 28, 1997

NC State Engineers' Software Speeds Up Motor Application Development

Electric motors are the heart of today's industry. We rely so much on their use that when a motor breaks down, a company's productivity and reliability are heavily affected, sometimes costing millions of dollars in the form of repairs and lost profits. Choosing the correct types of motors to use and properly maintaining them are essential steps in the process of creating an efficient, cost-effective, and failure-free production environment. The process can be greatly simplified and accelerated if an appropriate computer simulation tool is available.

Dr. Mo-Yuen Chow, associate professor of electrical and computer engineering, and his FasPro Research Group at North Carolina State University have designed a software package, MotorSIM, which makes modeling and simulation of electric motors much easier. This computer program allows engineers and designers to use pre-programmed building blocks to study the behavior of many types of electric motors. These modules can be customized quickly and easily to simulate various operating and fault conditions.

"MotorSIM is a computer software package that provides a framework for in-depth simulation of motor dynamics," says Chow. "It meets the needs of today's industrial competitive environments, offering cost effectiveness, flexibility, safety, accuracy and time efficiency for fast prototyping platforms."

Computer-aided design of motor systems is not a new idea. However, generally the engineers and designers must develop the programming code for each individual component of the motor--a time-consuming effort that requires close attention to detail. Any inaccuracy in the programming code can result in degraded performance in real-world motor applications. Because MotorSIM eliminates the need for the motor engineers and designers to develop computer code, the computer simulation and verification time can be reduced to a minimum, and the risk of inaccurate models can be greatly reduced.

MotorSIM has an extensive database of pre-programmed motor components, and its user interface is menu- and "wizard"-driven for easy data entry. Engineers and designers can click and drag icons representing various motor components into a computer model, run simulations and change parameters to easily find the best scenario for the application at hand.

Currently, MotorSIM can simulate four types of motors: three-phase induction, single-phase induction, universal and DC, including sub-modules such as electrical, mechanical, thermal, saturation, vibration and load.

"MotorSIM can simulate a motor's performance with constant and variable loads, temperature effects, magnetic and mechanical vibrations, and other relevant factors," says Chow. "It also enables fault injections, such as friction and winding faults and voltage surges and dips."

Working with his research assistants, Sinan Altug, Gregory Goddu, Bo Li, Alberico Menozzi and Jinxiang Zhu, Chow has formed the FasPro Research Group and developed the MotorSIM software package. The research is supported in part by the National Science Foundation and NC State University's Electric Power Research Center. In addition, Square D Company has loaned several pieces of equipment for the MotorSIM project.

The FasPro Research Group is also in the process of creating a computerized motor fault detector that provides a non-invasive technique to predict and analyze faults without having to dismantle the motor. The fault detector uses sensors to read and measure the motor's "vital signs"--current, speed, temperature and other signals--in much the same way a medical doctor uses a patient's vital signs to diagnose an illness. Based on the motor readings, the motor fault detection software, which includes a combination of fuzzy logic and neural network technologies, determines the motor's health and decides whether corrective action needs to be taken.

"Artificial neural networks provide precise solutions to problems while fuzzy logic is easy to implement and uses heuristic reasoning. The combined synergy of fuzzy logic and neural networks can provide accurate solutions for different operating conditions. It is an inexpensive and safe way of performing accurate fault detection," says Chow.

In his laboratory at the Electric Power Research Center at NC State University, Chow has proven the success of his fuzzy/neural fault detectors and is working to refine the software program for market production.

Chow anticipates having the MotorSIM software on the market this year and predicts that the Fuzzy/Neural Fault Detection software will be available in two to three years.

/ News Index / News Archives Index /