Disk Head Positioning Controller Reducing Resonance via Dynamic Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional magnetic disk devices face challenges in high-speed head positioning due to mechanical resonance, leading to vibration and noise, especially when trying to move the head over long distances, as existing techniques require extensive memory storage for feedforward control inputs and target position commands, and simple velocity feedback systems result in overshoot and resonance excitation.
Innovation Solution
A magnetic disk device with a positioning controller that includes a feedback controller and a model controller using a preset control equation model to generate feedforward control commands and target position commands, based on the differential value of target velocity corresponding to the remaining distance, allowing smooth acceleration and deceleration without exciting mechanical resonance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a simple velocity feedback control system is used to make model velocity follow target velocity, then the control system is simple, but the model velocity cannot follow the target velocity when target velocity is increased, causing overshoot
Solution Approach 1:
The patent introduces a velocity feedback mechanism where the velocity error (difference between model velocity and target velocity) is multiplied by a gain to generate a feedback control command. This feedback is combined with feedforward control commands to produce the final actuator control signal, enabling the model velocity to accurately follow the target velocity curve without overshoot.
Solution Approach 2:
The patent dynamically adjusts control parameters including velocity feedback gain, acceleration feedback gain, and jerk feedback gain based on the current state (velocity error, acceleration error, jerk error). This adaptive parameter adjustment allows the control system to maintain optimal performance across different operating conditions and velocity ranges.
2Measurement precision
If velocity feedback gain is set to high value to improve following capability, then model velocity can follow target velocity better, but velocity is abruptly switched to deceleration causing feedforward control command with abrupt switching, exciting mechanical resonance
Solution Approach 1:
The patent makes the control system dynamic by continuously adjusting control parameters (velocity feedback gain, acceleration feedback gain, jerk feedback gain) based on real-time error signals. This dynamic adaptation allows smooth transition between acceleration and deceleration phases, avoiding abrupt velocity switching that would excite mechanical resonance while maintaining accurate velocity following.
Solution Approach 2:
The patent introduces acceleration feedback and jerk feedback as intermediary control elements between velocity error and the final control command. These intermediary feedback mechanisms smooth the control signal transitions, preventing abrupt switching while maintaining velocity following accuracy, thereby suppressing mechanical resonance excitation.
3Ease of operation
If feedforward control input and target position command for all seek distances are held in table, then positioning control can be implemented, but memory capacity of microprocessor is exceeded
Solution Approach 1:
The patent segments the control approach by separating feedforward control (pre-calculated optimal velocity profile) from feedback control (real-time error correction). This segmentation allows the system to use minimal stored data (only the feedforward velocity profile) while generating control commands through real-time calculation, dramatically reducing memory requirements compared to storing complete control solutions for all possible seek distances.
Solution Approach 2:
The patent employs dynamic parameter adjustment where control gains (velocity feedback gain, acceleration feedback gain, jerk feedback gain) are continuously adapted based on current error signals. This dynamic approach enables the system to handle all seek distances with a unified control algorithm rather than requiring pre-stored solutions for each distance, reducing memory requirements while maintaining control accuracy.
Data Source
AI summary
A feedback controller performs a feedback control based on an error between a target position and an actual position of a head. A model controller obtains a control command, a model position to be output as a command for the target position to the feedback controller, and a model velocity for the head to follow a target velocity, by using a preset control equation model, and obtains the control command and the model position with the target position as an input, based on a differential value of the target velocity corresponding to a remaining distance from the model position to the target position.