Hybrid Electric Ship Demand Load Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current demand load forecasting methods for hybrid electric ships are inadequate due to their reliance on known demand loads, which fail to account for uncertain factors during navigation, leading to inaccurate predictions and inefficiencies in energy management.
Innovation Solution
A method employing a least squares support vector machine (LS-SVM) for working condition classification, combined with a Markov chain forecasting method and a radial basis function neural network optimized by a genetic algorithm, to predict demand loads based on historical data and real-time characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If time series method is used for demand load forecasting, then the method can capture temporal patterns, but the process of selecting embedding dimension and time delay becomes cumbersome and complex
Solution Approach 1:
The patent transforms the complex parameter selection problem into an automatic optimization process by using genetic algorithms to search for optimal embedding dimensions and time delays, converting manual parameter tuning into an automated evolutionary search process that finds optimal parameters without human intervention
Solution Approach 2:
The system performs self-optimization by automatically selecting its own parameters through the genetic algorithm framework, where the model evaluates different parameter combinations and iteratively improves its configuration without external guidance, making the system self-configuring and adaptive
2Measurement precision
If neural network method or support vector machine is used for demand load forecasting, then the model can learn complex patterns, but the training speed becomes slow and large amounts of data are required
Solution Approach 1:
The patent divides the forecasting problem into multiple segments by applying different forecasting models to different working condition types (stable vs. fast-changing conditions), allowing each segment to be optimized independently with appropriate model complexity rather than using a single large model for all conditions
Solution Approach 2:
The system dynamically selects the appropriate forecasting model based on the current working condition type, switching between simpler models for stable conditions and more complex models for fast-changing conditions, making the system adaptive and efficient rather than statically using a fixed model
3Ease of manufacture
If known demand loads are used as basis for energy management strategies, then the strategies can be formulated in advance, but they fail when uncertain factors occur during actual navigation
Solution Approach 1:
The patent prepares multiple forecasting models and working condition classification frameworks in advance, so when uncertain factors occur during navigation, the system can immediately switch to appropriate pre-prepared models rather than formulating strategies from scratch, enabling rapid adaptation to changing conditions
Solution Approach 2:
The system continuously monitors actual working conditions and compares them with forecasted demand loads, using this feedback to update and refine the forecasting models in real-time, ensuring the energy management strategies remain reliable even as navigation conditions change and uncertain factors emerge
Data Source
AI summary
The disclosed is a method for forecasting a demand load during navigation of a ship based on working condition classification. The method of the disclosure comprises: training historical data values by employing a least squares support vector machine to obtain a classification plane, classifying working conditions during the navigation into a fast-changing working condition and a stable working condition, and determining in real time a working condition type of the ship in an online stage. A method for forecasting a demand load by means of a Markov chain is employed for the stable working condition, and a method for forecasting a demand load by means of a radial basis function neural network optimized by a genetic algorithm is employed for the fast-changing working condition. A desirable forecasting effect can be achieved by selecting forecast models suitable for either type of working condition.