Hydraulic controller nets best-paper win

Mondeep Duarah and Vijay Jayabalan of Captronic Systems, Bangalore, India, won Best Paper in the control category at NIWeek 2007 with their article entitled “Digital Electro-Hydraulic Servo Controller.”

By Control Engineering Staff August 23, 2007

Austin, TX —Mondeep Duarah and Vijay Jayabalan of Captronic Systems, Bangalore, India, won Best Paper in the control category at NIWeek 2007 with their article entitled “ Digital Electro-Hydraulic Servo Controller .” The paper describes a PC-based digital controller they developed using LabView and PXI/Compact PCI hardware from National Instruments (NI).

“Although electrical motors are used in many motion control systems, electro-hydraulic systems often address applications requiring either very high force or wide bandwidth more efficiently,” they point out in the paper. “In general, electro-hydraulic systems better serve applications with control bandwidths greater than about 20 Hz, or control power greater than about 15 kW.”

Such servo-hydraulic systems are mechanically “stiffer,” producing higher machine frame resonant frequencies for a given power level, higher loop gain, and improved dynamic performance. The driving fluid effectively acts as a cooling medium, carrying heat away from the actuator and flow control components, making such systems run cooler.

“Unfortunately,” the authors continue, “hydraulic systems also exhibit several inherent non-linear effects which can complicate the control problem.”

The fact that the vast majority of electronic closed-loop controllers are based on simple analog circuit designs makes it difficult to deal with these non-linear effects. Control designers have traditionally chosen these designs because they offer robust, low cost implementations of the well known PID control strategy. This approach works well in systems with simple topology and limited bandwidth.

The growing use of complex control strategies and complicated mechatronic architectures, however, coupled with the need to support enhanced features such as data logging and digital communications, has led to increased interest in digital processors for control of hydraulic servo-systems.

The authors list some of the benefits of going to a PC-based digital controller:

Improved immunity to component tolerances, thermal drift, and aging;

Improved noise immunity;

Ability to modify and store control parameters;

Ability to easily implement digital communications;

System fault monitoring and diagnostic capabilities;

Data logging capability;

Automated calibration;

Diagnostic monitoring identify mechanical vibrations and predict failure modes;

High-order digital filters to suppress resonant modes; and

Lower overall system cost.

The test system that the authors built uses an electro-pneumatic actuator load a test specimen according to a test-waveform created by a PXI-based signal generator capable of producing sine, triangle, square, or combination waveforms.

To control loads on the test specimen strain-gauge load cells monitor the actual load. The resulting electrical signals go through a National Instruments SCXI-1520 signal conditioner and compared to the original test waveform. The error signal is then amplified and used to control a servo valve, which controls oil flow rates from a high pressure pump to the actuator, driving the piston to modify the force on the test specimen.

Control is achieved by using a software implementation of a PID controller NI PID Control Toolkit. The PID controller uses the position form of the PID algorithm. PID controller is capable of loading the system defined by the wave form. The PID Control toolkit provided resources for shunt calibration, offset nulling, and programmable excitation.

Unfortunately, the nonlinear behavior of the system forces conservative loop gains and the gain scheduling feature of the PID Control Toolkit helped implement them. The controller was tuned at multiple operating points to take care of the systems nonlinearity.

The authors say that the digital servo controller developed is highly flexible, rugged, and cost effective. Tests show that it can achieve an average accuracy of 2 A

Other Control Engineering NIWeek 2007 coverage includes software advancements.

C.G. Masi , senior editor, Control Engineering