Power Window Performance Prediction Using LSTM Component Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for predicting the performance of a power window in vehicles require repeated measurements with actual devices, especially when peripheral parts are replaced or durability changes, leading to inefficiencies and inaccuracies due to the lack of a predictive technique.
Innovation Solution
A deep learning model, specifically using a Long Short Term Memory (LSTM) network, is trained with data including slide resistance, stroke distance, weight, torque, and durability of the power window to predict the performance, reducing the need for repeated measurements and providing high-accuracy results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If measurement methods using actual measuring devices are used to predict power window performance, then measurement accuracy can be ensured, but time consumption increases significantly due to repeated measurements whenever peripheral parts are replaced or durability changes
Solution Approach 1:
The system performs preliminary measurements and stores performance data (operating current, operating time, slide resistance, torque, durability) in a database during the design and testing phase. This pre-collected data is then used by the prediction model to forecast performance without requiring repeated actual measurements, thus resolving the contradiction between measurement accuracy and time consumption.
Solution Approach 2:
The system creates a virtual copy of the power window system through a prediction model that replicates the behavior of the actual power window. This digital twin or virtual model uses stored historical data to simulate and predict performance characteristics, eliminating the need for physical repeated measurements while maintaining prediction accuracy.
2Reliability
If repeated measurements are performed whenever peripheral parts (glass run, glass, motor) are replaced, then updated performance data can be obtained, but productivity decreases due to the repetitive nature of the process
Solution Approach 1:
The system implements a feedback mechanism where the prediction model continuously compares predicted performance with actual measured data from the database. When peripheral parts are replaced, the system automatically updates the database with new component parameters (slide resistance, torque, weight) and re-runs predictions, providing continuous feedback without requiring manual repeated measurements, thus maintaining reliability while improving productivity.
Solution Approach 2:
The prediction system serves itself by automatically updating performance predictions when component changes are detected. The system autonomously retrieves updated component data from the database, re-trains or re-executes the prediction model, and generates updated performance forecasts without requiring external intervention or repeated physical measurements, thereby enhancing both productivity and data reliability.
3Measurement precision
If individual measurements are conducted for each vehicle model to account for different peripheral part specifications, then model-specific accuracy is achieved, but device complexity increases
Solution Approach 1:
The prediction system is designed as a universal multi-functional platform that can handle multiple vehicle models and different peripheral part configurations through a single integrated system. The database stores diverse component data across different models, and the prediction model automatically adapts to different vehicle types based on input parameters, eliminating the need for separate measurement systems for each model while maintaining model-specific accuracy.
Solution Approach 2:
The system manages model-specific variations by changing parameters (slide resistance, torque, weight, durability values) in the database rather than changing the physical measurement system. The prediction model dynamically adjusts its calculations based on the specific parameters of each vehicle model and peripheral part configuration, achieving model-specific accuracy through parameter variation rather than system complexity.
Data Source
AI summary
The present disclosure relates to an apparatus for predicting performance of a power window and a method thereof. The apparatus for predicting performance of a power window may include a memory storage that stores a deep learning model and trained updates thereto and a controller that trains the deep learning model to predict the performance of the power window using a slide resistance of a glass run, a stroke distance of a door glass, a weight of the door glass, a torque of a motor, and a durability of the power window. The system may then predict performance of a target power window based on the deep learning model on which training has been performed.