DC/DC Converter Fault Diagnosis Using Sparrow Search and DBN
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fault diagnosis methods for DC/DC converters, particularly for dual-active-bridge converters, face challenges in establishing accurate mathematical models and suffer from low accuracy due to shallow learning techniques and optimization algorithms that often fall into local optimal solutions, leading to long calculation times and low diagnosis accuracy.
Innovation Solution
An improved sparrow search algorithm is used to enhance the global optimization ability, combined with a Levy flight strategy, to optimize the number of hidden-layer units in a deep belief network, thereby improving the fault diagnosis accuracy by preventing overfitting and local optimal values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a deep belief network is used for fault diagnosis, then feature extraction ability is improved, but the risk of falling into local optimal solutions increases
Solution Approach 1:
The sparrow search algorithm is introduced as an intermediary optimization method between the deep belief network and the fault diagnosis task. It serves as a mediator that optimizes network parameters without requiring extensive manual experimentation, thereby maintaining high diagnostic accuracy while reducing the risk of local optimal solutions through its nature-inspired search mechanism.
Solution Approach 2:
The sparrow search algorithm dynamically adjusts optimization parameters during the fault diagnosis process. By changing search strategies and parameter settings based on iterative progress, it adapts to avoid local optima while maintaining the deep belief network's high feature extraction capability for accurate fault diagnosis.
2Productivity
If optimization algorithms are used to optimize network parameters, then parameter optimization efficiency is improved, but calculation time increases
Solution Approach 1:
The sparrow search algorithm employs periodic action principles by iteratively adjusting optimization parameters in cycles. It periodically updates search strategies and parameter configurations, allowing efficient parameter optimization while controlling calculation time through structured iterative processes rather than continuous exhaustive search.
3Device complexity
If shallow learning methods are used, then model complexity is reduced, but fault diagnosis accuracy decreases
Solution Approach 1:
The patent replaces manual mechanical experimentation for parameter optimization with an automated sparrow search algorithm. This substitution maintains the deep belief network's complex structure needed for high accuracy while eliminating the need for extensive manual parameter tuning, thus preserving diagnostic accuracy without proportionally increasing overall system complexity.
Data Source
AI summary
A DC/DC converter fault diagnosis method based on an improved sparrow search algorithm, includes: establishing an simulation module of the converter, selecting a leakage inductance current of a transformer as a diagnosis signal, and collecting diagnosis signal samples under OC faults of different power switching devices of the converter as a sample set; improving a global search ability of a sparrow search algorithm by using a Levy flight strategy; dividing the sample set into a training set and a test set, preliminarily establishing an architecture of a deep belief network, and initializing network parameters; optimizing a quantity of hidden-layer units of the deep belief network by using an improved sparrow search algorithm, to obtain a best quantity of hidden-layer units of the deep belief network; and training an optimized deep belief network obtained based on the improved sparrow search algorithm, and obtaining a fault diagnosis result based on a trained network.