High-Speed Train Speed Prediction Control With Forecast Error Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-speed trains present challenges due to their non-linear and uncertain dynamics, making it difficult to accurately model and control their speed, especially with traditional methods like PID control and neural networks, which require extensive data and are not suitable for real-time responsive systems.
Innovation Solution
A method and system for state feedback predictive control based on forecast error, using a speed prediction model to correct prediction speeds and calculate control forces, enhancing adaptability and accuracy while simplifying calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If traditional dynamic equations are used to model the high-speed train, then the model structure is simple, but the modeling accuracy deteriorates due to high nonlinearity and uncertainty
Solution Approach 1:
The patent introduces an intermediary mechanism by combining traditional dynamic equations with a neural network compensator. The neural network acts as an intermediary that captures the nonlinear and uncertain dynamics that the traditional equations cannot represent, thereby improving modeling accuracy while maintaining the structural simplicity of the traditional model.
Solution Approach 2:
The patent creates a composite modeling approach by integrating two different modeling paradigms: traditional physics-based dynamic equations and data-driven neural networks. This composite model leverages the strengths of both approaches - the interpretability and simplicity of physics-based models and the accuracy of neural networks in capturing complex nonlinear behaviors.
2Measurement precision
If neural network is used to model the high-speed train, then the modeling accuracy improves, but the time cost increases and it may fall into local optimum
Solution Approach 1:
The patent segments the modeling task into two parts: a traditional dynamic equation model that provides a baseline structure and a neural network component that only needs to learn the residual nonlinear and uncertain dynamics. This segmentation reduces the complexity and data requirements for neural network training compared to training a complete neural network model from scratch.
Solution Approach 2:
The patent performs preliminary action by using traditional dynamic equations to establish the basic model structure before applying neural network compensation. This preliminary modeling step reduces the burden on the neural network to learn all dynamics from scratch, thereby reducing training time and avoiding local optima issues.
3Ease of operation
If PID control algorithm is used to control the high-speed train, then the control implementation is simple, but the control performance deteriorates due to reliance on artificial parameter adjustment
Solution Approach 1:
The patent implements a state feedback mechanism where the control input is determined by the current system state and the desired trajectory. This feedback structure automatically adapts the control parameters based on the current operating conditions, eliminating the need for manual parameter adjustment while maintaining simple control implementation through a unified feedback law.
Solution Approach 2:
The patent transforms the static PID parameter adjustment into a dynamic state-feedback control law where control parameters are automatically adjusted based on the current system state. This dynamic approach maintains implementation simplicity while significantly improving adaptability to changing operating conditions.
4Measurement precision
If generalized prediction control algorithm is used for speed tracking, then the control accuracy improves, but the calculation complexity increases making it unsuitable for real-time control
Solution Approach 1:
The patent extracts and separates the complex calculation components from the prediction control algorithm, retaining only the essential predictive functionality. By focusing on speed tracking rather than full-state prediction and using a simplified control law, the patent achieves real-time applicability while maintaining adequate tracking accuracy.
Solution Approach 2:
The patent applies partial action by implementing prediction control only for the speed variable rather than for all system states. This partial implementation reduces calculation complexity significantly while still achieving the primary control objective of speed tracking, making it suitable for real-time control systems.
5Adaptability or versatility
If fuzzy control is used for automatic parking and running process control, then the control flexibility improves, but the calculation difficulty increases due to experience-based fuzzy rules and membership functions
Solution Approach 1:
The patent replaces the complex fuzzy logic system with a more efficient state-feedback control mechanism. Instead of using experience-based fuzzy rules and membership functions that require complex calculations, the patent uses a direct state-feedback law that achieves similar flexibility and adaptability with simpler computations suitable for real-time control.
Data Source
AI summary
The invention discloses a method and a system for state feedback predictive control of a high-speed train based on a forecast error. The method comprises: obtaining a speed prediction model of a high-speed trainy^k+p=CApxk+∑i=1pCAi-1Buk+p-i;predicting speeds of the train at times k and k+p according to the speed prediction model; obtaining an actual speed output value of the train; determining a speed prediction error at time k according to the prediction speed of the train and the actual speed output value of the train; correcting a prediction speed of the train at time k+p, according to the speed prediction error, to obtain a corrected prediction speed of the train; calculating a control force uk of the train according to uk=λ−1(p)[yk+pr−yk−Kxk+ŷk]; and applying a control force to the train based on the control force uk. The disclosure may have features of strong adaptability and easy calculation, and achieve high-precision tracking of train speed.