Generator Load Estimation Using Voltage Current Feature Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the type of load connected to a power generator require vast training data for neural networks or extensive frequency analysis, making them complex and inefficient for accurate load estimation.
Innovation Solution
A load estimating device with a measuring circuit to measure voltage and current, a calculating unit to calculate feature amounts, and a storage device to store and match these values for accurate load identification, allowing for simpler and more precise load estimation, including a registration mode for new loads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If neural network learning is used to determine operation state, then determination accuracy is improved, but training data requirements and system complexity increase
Solution Approach 1:
The patent extracts only the essential features needed for load estimation (voltage and current measurements) rather than using complex neural networks. The calculating unit computes specific feature amounts from these measurements, and the load estimating unit compares them against stored reference values, achieving accurate load identification with minimal computational complexity.
Solution Approach 2:
The patent uses simple, inexpensive measurement and calculation components instead of expensive neural network systems. The measuring circuit, calculating unit, and storage device form a cost-effective solution that provides sufficient accuracy for load estimation without requiring vast training datasets or complex processing infrastructure.
2Measurement precision
If frequency analysis is used to determine operation state, then determination accuracy is improved, but calculation requirements and processing time increase
Solution Approach 1:
The patent extracts only the necessary voltage and current measurements from the electrical signal, avoiding comprehensive frequency analysis. The calculating unit computes feature amounts directly from these measurements, enabling rapid load estimation without the heavy computational burden of full spectrum analysis.
3Reliability
If comprehensive frequency analysis is performed, then operation state determination is improved, but system resource consumption increases
Solution Approach 1:
The patent employs energy-efficient measurement and calculation components that consume minimal computational resources. The system achieves reliable load identification through simple voltage and current measurements with basic feature amount calculations, avoiding the high energy consumption associated with comprehensive frequency analysis and neural network processing.
Data Source
AI summary
A load estimating device measures a voltage and a current supplied to a load connected with a generator, calculates a feature amount of the load, senses a remaining amount of fuel, outputs a time during which the load is continuously operable. The device estimates what the load connected with the generator is, based on the calculated feature amount and the feature amounts stored in a storage, and determines the time during which the estimated load is continuously operable, based on a power consumption of the estimated load, and the remaining amount of fuel. The device has a load registration mode for causing the storage to store therein a feature amount of a new load that is not stored in a storage.