This proprietary feature has no peer in the motion control industry. WorldServo’s engineers have gone one large step closer toward removing the need for concern over the motor-load inertia match by perfecting a new adaptive control technique, aptly named Inertia Matching Technology.
Concern over the inertial match between motors and loads is required for two reasons:
  1. A high inertia mismatch can cause a low frequency resonance to occur limiting the usable gain of the servo system, and
  2. The "under-powered" nature of the axis when confronted with large inertia loads can lead to excessive overshoot or complete instability unless the gains (especially the integrator (I gain) is lowered drastically).

Resonance control torque filtering is used to mitigate the first problem, either directly with a digital low-pass filter or by adjusting the response of the torque controller.
The second problem has troubled servo system designers and users for years. Here’s how this problem normally manifests itself:
The position/velocity compensator’s integrator is used to drive steady-state errors to zero in a servo system. In general, this is a good thing. The higher the integrator gain, the faster the errors are driven to zero, increasing the dynamic stiffness of an axis. Unfortunately, until now, using the integrator always leads to overshoot when responding to disturbances because of the very nature of the integrator. The higher the integrator, gain the larger the overshoot. Lowering the integrator gain lowers the amplitude of the overshoot but increases the duration. If you have an axis that has less than perfect mechanics and/or the inertia mismatch becomes significant, you are likely to have problems. Eventually, it becomes impossible to keep the system stable for even the smallest disturbance.

The Historical Solution
Ineffective and/or highly compromised solutions to this problem in the past have included: turning off the integrator until the end of a move, clamping the maximum value of the integrator, providing a window in which the integrator operates, etc. All of these solutions are of little utility and many of them actually degrade performance. The common, brute-force way of dealing with this problem is to over-size the motor (select a motor with torque/force well beyond what the application requires), keep the feedback gains moderate and use very little or no integrator gain.
This over-sizing of the motor is usually effective Because the motor is effectively "unloaded" by this technique, the integrator is seldom required for accuracy, so the I-gain can be kept low or turned off. The problem is, of course, that this over-sizing solution is both expensive and, in many applications, the motor is just too large. However, the success of this expensive over-sizing solution has led to a well-accepted inertia matching rule-of-thumb. This guideline is promoted by many servo system salesmen because they know if they get the customer to engineer in this way they will get fewer support calls (and more revenue from larger systems).

Enter the SSt's IMT
So how does the SST servo system solve this problem? The SST’s IMT feature uses neural/fuzzy adaptive control techniques designed using thousands of hours of simulation, and tested rigorously on a wide variety of axes. The IMT eliminates overshoot caused by large disturbances while maintaining high stiffness.
It does this by simultaneously modulating the gains in the compensator during disruptive events. Because this technique is highly proprietary, we can’t tell you exactly how this works, but we can show you how well it works. Scope shots from the Real-time Monitor Port below display the response of an SSt servo system to an instantaneous step change in commanded position. The top trace is a well-tuned SST system with a reasonable integrator value. The lower trace, to those with servo experience, might look as if the integrator gain has been set to zero—in fact, it was increased by forty percent! Notice that the response is actually faster with no overshoot!