Motor Brake Failure Prediction Using Normal-State Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods fail to accurately predict motor brake failure in advance without relying on sensors, as data variability due to temperature and motor type complicates threshold setting, and sensor-based predictions increase costs.
Innovation Solution
A machine learning device that observes and learns from normal state data, including gravity load torque, mechanical friction torque, and brake reaction time, to predict motor brake failure without the need for additional sensors, by automatically determining failure thresholds based on environmental conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor-based measurement methods are used to improve brake failure prediction accuracy, then prediction accuracy is improved, but system cost increases
Solution Approach 1:
The motor controller utilizes existing operational data (current, voltage, speed, acceleration) that is already being collected for motor control purposes. The brake failure prediction function is implemented through software algorithms that process this existing data without requiring additional sensors or measurement devices, thereby achieving improved prediction accuracy while avoiding increased system cost
Solution Approach 2:
The motor controller performs multiple functions: it controls motor operation and simultaneously predicts brake failure using the same operational data. This multi-functionality allows the system to gain brake monitoring capabilities without adding dedicated sensors, as the existing control system is leveraged for dual purposes
2Ease of operation
If fixed threshold values are used for brake failure detection, then detection simplicity is improved, but prediction accuracy deteriorates due to data variability from temperature and motor type
Solution Approach 1:
The system dynamically adjusts the threshold values for brake failure detection based on operating conditions such as temperature and motor type. Instead of using fixed thresholds, the thresholds are adapted in real-time to match current environmental and operational parameters, thereby maintaining detection simplicity while significantly improving prediction accuracy across varying conditions
Solution Approach 2:
The system changes the threshold parameters according to the specific motor type and operating temperature. Different threshold values are applied for different motor models and temperature ranges, allowing the detection system to remain simple in operation while accurately accounting for data variability caused by different operating conditions
Data Source
AI summary
A failure prediction device is provided with a machine learning device configured to learn the state of a brake of a motor with respect to data on the brake. The machine learning device observes brake operating state data indicative of an operating state of the brake when the brake is in a normal state, as state variables representative of a current environmental state, and uses the observed state variables to learn a distribution of the state variables with the brake in the normal state.