Model-Based Compensation of Dynamic Errors in Measuring Machines and Machine Tools

The demand for fast automated inspection of all manufactured components in industrial production is rising due to increasing quality demands and high-speed machining technologies. Modern surface metrology sensors are capable of implementing these tasks if they can be embedded in fast and accurate measuring machines. However, the common approach to achieve high positioning accuracy in such machines is abiding by very conservative limits in acceleration, and thus, velocity. We propose an active control system including a model-based compensation system for dynamic errors to increase acceleration and velocity. A novel optical sensor permitting the direct measurement of the actual tool-center-point (TCP) position is used for system identification of the required dynamic model. The position of a LED point source is detected with high subpixel precision by replicating the spot several times on the camera chip and averaging over the center of gravities of all spots. A model for the largest dynamic error component of a small multiaxis measuring machine is derived and the sensor is used to identify the model parameters. A model-based feedforward and feedback control system reducing the dynamic tracking error is designed. This control systems allows machine motions with both high accuracy and large velocity. The achieved performance is evaluated using TCP tracking error measurements obtained with the novel optical sensor. This model-based compensation system can also be applied to machine tools.