Optimal Solution Calculation Using Previous-Solution Initialization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for solving optimization problems in model predictive control, such as analytically solving mathematical expressions and gradient methods, face difficulties with restrictions on control targets and evaluation functions, leading to increased calculation time due to the need for random initial solutions in search-based methods.
Innovation Solution
An optimal solution calculation device that generates initial solutions using previous optimal solutions, allowing for parallel processing and adjusting the number of search repetitions based on evaluation function values to reduce the number of searches and calculation time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If random initial solutions are used in search-based methods, then the method can solve difficult optimization problems, but the calculation time increases due to the need for extensive searches
Solution Approach 1:
The patent applies preliminary action by generating initial solutions from previous optimal solutions before starting the search process. This pre-positioning of initial solutions in promising areas of the search space eliminates the need for random initialization, reducing the number of searches required while maintaining the ability to solve difficult optimization problems.
Solution Approach 2:
The patent implements feedback by using the optimal solution obtained in one time step as the initial solution for the next time step. This feedback mechanism allows the search process to leverage previously acquired knowledge, significantly reducing calculation time while maintaining reliability in solving optimization problems with time-varying parameters.
2Measurement precision
If the number of search repetitions is increased to avoid local solutions, then the optimal solution accuracy improves, but the calculation time increases
Solution Approach 1:
By pre-generating initial solutions from previous optimal solutions, the patent reduces the number of search repetitions needed to achieve accurate optimal solutions. The initial solutions are already positioned in promising areas, so fewer iterations are required to converge to the optimal solution, maintaining accuracy while reducing calculation time.
Solution Approach 2:
The patent dynamically adjusts the search process by using time-varying initial solutions that adapt to changing parameters. This dynamic approach allows the system to maintain high solution accuracy with fewer repetitions by continuously leveraging the most recent optimal solution information, rather than using fixed random initial solutions.
3Loss of time
If analytical methods are used to solve the evaluation function, then the calculation time is reduced, but the control target is restricted
Solution Approach 1:
The patent segments the solution process into two parts: using analytical methods for the evaluation function calculation (to reduce calculation time) and using search-based methods with improved initial solutions for finding optimal control inputs (to maintain versatility). This segmentation allows the system to handle diverse control targets while maintaining efficient calculation.
Solution Approach 2:
The patent introduces an intermediary mechanism (the improved initial solution generation from previous optimal solutions) that bridges analytical methods and search-based methods. This intermediary allows the system to use fast analytical evaluation while still achieving versatile optimal control for various control targets through the guided search process.
Data Source
AI summary
An optimal solution calculation device repeatedly calculates an optimal solution that minimizes an evaluation function having a time-varying parameter at regular intervals. The optimal solution calculation device includes: an initial solution generation unit that generates an initial solution for providing an initial value for a search process to search for the optimal solution; and an optimal solution search unit that calculates the optimal solution through the search process using the initial solution. The initial solution generation unit generates the initial solution using a previous optimal solution calculated by the optimal solution search unit.