Multi-Signal Failure Prediction for Hard Disk Drives
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing failure prediction methods for electronic devices, such as hard disk drives, face challenges in accurately identifying failed components due to reliance on single detection signals like electric current, which can be affected by multiple factors, leading to incorrect predictions.
Innovation Solution
A failure prediction system that utilizes multiple state signals, including vibration and current signals, synchronizes and analyzes these signals to calculate feature values, allowing for accurate failure prediction by comparing them against reference values, and includes a signal processing circuit with a phase processing part, signal conversion part, and failure prediction part.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If only electric current is used as detection signal, then the detection system is simple, but failure prediction accuracy is low
Solution Approach 1:
The patent combines multiple detection signals (vibration signal from acceleration sensor and current signal from current sensor) into a unified failure prediction system. The vibration signal detection unit and current signal detection unit work together, with their outputs fed into the same signal processing circuit and failure prediction part, enabling accurate identification of failed portions through multi-parameter analysis.
Solution Approach 2:
The detection system is segmented into independent functional units: vibration signal detection unit with acceleration sensor, current signal detection unit with current sensor, signal processing circuit with phase processing and signal analysis parts, and failure prediction part. Each unit processes specific signals independently before integration, allowing modular complexity management while achieving high prediction accuracy.
2Measurement precision
If multiple detection signals are used, then failure prediction accuracy is improved, but the system complexity increases
Solution Approach 1:
The signal processing circuit serves multiple functions: it processes both vibration signals and current signals, performs phase processing on both types of signals, calculates feature values from both signal types, and integrates their analysis results. This multi-functional design handles multiple detection signals uniformly, reducing overall system complexity despite increased detection capabilities.
Solution Approach 2:
The signal processing circuit acts as an intermediary between the multiple detection signals and the failure prediction part. It receives both vibration and current signals, synchronizes their phases, extracts feature values from each, and prepares integrated analysis results for the failure prediction part, simplifying the interface between diverse sensors and the prediction algorithm.
3Measurement precision
If single detection signal is used, then the system is easy to operate, but incorrect predictions occur due to inability to identify abnormal portion
Solution Approach 1:
The system provides feedback by analyzing both vibration and current signals to determine which specific component has failed. The failure prediction part receives integrated analysis results from both signal types and outputs predictions about specific failed portions (spindle motor, magnetic head, slider, head arm, or voice coil motor), enabling accurate diagnosis without increasing operational complexity for the user.
Data Source
AI summary
A failure prediction system for performing failure prediction to a monitoring target device by detecting a state, comprising: a state detection unit for detecting state signals of no smaller than two different kinds, and outputting a detection signal corresponding to each of the state signals; a phase processing part for synchronizing a plurality of the detection signals; a signal analysis part for calculating a feature value indicating a feature of the state for each of the detection signals from the phase processing part; and a failure prediction part for performing failure prediction of the monitoring target device for each of the feature values by comparing the feature value in question and a reference value set in advance.