Engine Model Training Patterns for High-Coverage Test Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current engine testing methods require repeated cycles of construction and evaluation, resulting in a significant amount of test man-hours needed to achieve a desired response, which is inefficient.
Innovation Solution
An engine model construction method that generates and corrects Chirp signals to maximize coverage of manipulated variables and their change rates, acquiring time series data through engine tests, and constructing a machine learning model using these patterns and data to predict controlled amounts, thereby reducing test man-hours.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If engine testing is performed using conventional methods with repeated construction and evaluation cycles, then the desired response can be achieved, but a significant amount of test man-hours is required
Solution Approach 1:
The patent applies preliminary action by constructing a virtual engine model before actual engine testing to predict engine behavior and optimize test patterns. The virtual model allows pre-evaluation of test conditions, so that when actual testing occurs, the test man-hours are significantly reduced because the testing can proceed directly with optimized parameters rather than requiring repeated construction and evaluation cycles
Solution Approach 2:
The patent creates a virtual engine model that copies the essential characteristics and behavior of the actual engine. This virtual copy can be tested repeatedly without consuming physical test resources or requiring reassembly of the actual engine, thereby achieving the desired response while minimizing test man-hours
Data Source
AI summary
An engine model construction method includes generating test patterns in which a plurality of manipulated variables used for an engine test are changed with time, correcting the test patterns based on first coverage of a first space of manipulated variables are allowed to take and second coverage of a second space of change rate values of the manipulated variables are allowed to take, acquiring pieces of time series data of operation amounts of the manipulated variables and controlled amounts with respect to the manipulated variables by performing an engine test using the corrected test patterns, and constructing a first engine model by performing machine learning on training data in which the corrected test patterns are adopted as input and the pieces of time series data are adopted as correct answers, by a processor.