Slow work

Engineers at Nottingham University aim to design a drive that overcomes sensorless designs' inability to operate at zero or low speeds.

With industrial partners

,

,

, they hope to develop drives for a variety of applications.

Controlling AC motors that can change speed and torque quickly is essential for advanced industrial applications, from machine tools to large drives such as those in paper mills.

However, they require a lot of instrumentation, such as position sensors mounted on the drive shaft, which affect the reliability of the system, as there are numerous things that can go wrong. The signal from the motor to the control can be distorted by electromagnetic interference and, in a factory setting, it is easy for a cable to be broken.

The group is trying to control the motor without measuring position or torque.

'Our "box of tricks" that does all the intelligent stuff is a power converter, which takes mains frequency fixed-voltage supply and converts it into a variable frequency voltage supply,' said project leader Mark Sumner. 'Three power cables connect it to the motor, with no feedback signals from the motor.'

While much work has been done on sensorless control in the past 20 years, Sumner said it breaks down at zero and low speed. 'It always fails,' he claimed. 'But the technique we've developed works very well throughout the range.

'Past techniques used a mathematical model with signals picked up from the machine but these signals disappear at low and zero speeds. The signals result from the rotation of the motor and, as the motor slows down, they disappear.

'We inject test signals and because a motor is not a perfect device and has small non-linearities within it, we can pick up these non-linearities and detect the position of the rotor at any time, under any operating conditions.'

Sumner has already developed a system that works for one type of machine, but it has problems. 'We can get this position control at zero speed, but the sacrifice is a lot of extra audible noise, because the size of the injection signal is quite significant,' he said. 'Our aim is to reduce the size of the injection signal so that we can get precision in the position control, without the noise.'

The team has received a government grant of more than £300,000 to fund the project for the next three years and Sumner said that by the end of that time they hope to have engineered and optimised a prototype for an industrial drive.

More specialist applications are being investigated, notably in automotive and aerospace sectors. 'In the automotive area, we are looking at electric power steering,' he said. 'Noise is a real problem here, because a small version of one of these motors is bolted to the steering wheel and any extra vibrations get amplified by the whole steering column, giving an annoying noise in the passenger compartment. This new type of drive could overcome that.'

The team is also working on the more-electric aircraft. 'All the flaps and rudders on an aeroplane are controlled by hydraulics,' he said. 'We are looking to try to replace them with electric motor drives that are lighter and, potentially, more efficient.

'If we can reduce the number of transducers that are required on a drive, we can make it more reliable, and more fault tolerant.'

Both of these applications will take a little longer to achieve. 'For automotive applications, we need a 12V machine, whereas the work we've done so far is on 400V machines,' Sumner explained. 'There are subtle changes, and we are not sure what the implications are yet. It is also cost driven — although we have to get rid of the noise, we can't have an expensive microprocessor control or a large number of transducers because it all adds to the cost. We want high performance but at the same time it needs to be low-cost. It's a typical engineering problem.'

The team is already working with industrial manufacturing company Control Techniques on its application to process control. Sumner expects to bring Smiths and TRW (whom he already collaborates with in other areas) into the project at a later date, when the aerospace and automotive applications are further advanced.

The fourth company, PEM, designs and makes a transducer that measures the rate of change of current needed for the power converter and which will have to be engineered for each application.

'Its sensitivity, bandwidth and physical design will be a challenge, so we need their expertise,' said Sumner.