Drives and Controls Magazine
AI software compensates for faults in motors
Published:  20 April, 2018

Two Canadian companies have joined forces to commercialise an AI (artificial intelligence) technology designed to detect, and compensate for, faults in electric motors. The technology allows motors to generate near-normal torque even when there are inverter or motor phase faults. It can be applied to permanent magnet synchronous motors and to brushless DC motors (three-phase or multi-phase).

The two companies – Alizem, which specialises in power electronics applications and software, and Opal-RT Technologies, which is an expert in the real-time simulation of electrical systems – have signed a Memorandum of Understanding (MoU) to commercialise the edge AI technology.

“Electric motors are everywhere and their associated sensors are not only helpful in controlling them at desired speed and position, but also in closely monitoring their state and detecting associated faults,” says Alizem president, Dr Marc Perron. “The proposed embedded software algorithms go one step further in being able to compensate those faults in real-time and have the electric motor generate near-normal torque in the presence of inverter or electric motor phase faults.”

This is achieved by comparing a data-based model of the motor continuously with the model of a normal motor, and compensating for any discrepancies using voltage commands. “These algorithms can dramatically help in critical safety situations – like having a bus stuck on a railroad crossing and still being able to move a few metres ahead and get in a safe situation – hence changing the impact of a fault from catastrophic to manageable,” Perron explains.

The technology, which takes the form of a patented software algorithm, is designed to be integrated into the firmware of motor drives. Before it is integrated, a customer’s system can be simulated on an Opal-RT platform to determine how it will behave in an actual application. The final result is delivered either as software files, or as a complete prototype consisting of software, a controller, the power electronics and the motor.

Opal-RT CEO Jean Belanger (left) with Alizem president Marc Perron

Opal-RT’s CEO, Jean Bélanger, reports that his company has recently tested the technology and, within a few weeks, had achieved “very satisfying" results. He points out that designing motor drive systems is a complex process that can be simplified by using simulation systems to converge rapidly on a design “that meets dynamic and energy-efficiency requirements”.

“Today, one can go a step further in designing intelligent electric motor systems that are robust to faults without having to reinvent the wheel,” he adds. The idea is to make use of Alizem’s many years of r&d experience. “In a context where talent is rare-to-find, this IP-reuse approach is a great way to innovate, while saving costs, risks and time-to-market,” Bélanger says.

The technology is expected to have applications in areas such as robotics, electric vehicles and aircraft. It is currently available for licensing by OEMs to integrate into their motor control systems. It is not yet available as off-the-shelf software.