Submarine Cable Path Planning With Adaptive Parallel FMM
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for designing submarine cable paths are laborious, subjective, and fail to achieve optimal balance between cost and risk due to their lack of scalability and precision, especially for ultra-long-distance routes, and the sequential nature of the Fast Marching Method (FMM) impedes direct parallelization.
Innovation Solution
An Adaptive Parallel FMM (APFMM) is proposed, utilizing adaptive domain decomposition and multi-resolution analysis to decompose the target area into sub-domains and execute the FMM in parallel, iteratively refining resolution to achieve high-precision, ultra-long-distance submarine cable path planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual expert knowledge methods are used for submarine cable path planning, then subjective feasibility checks can be performed, but the process is laborious and fails to achieve optimal balance between cost and risk
Solution Approach 1:
The patent replaces manual expert knowledge methods with an automated computational system based on the Fast Marching Method (FMM). The system uses numerical algorithms to automatically calculate optimal submarine cable paths by solving the Eikonal equation, substituting human expert judgment with systematic computational processing that objectively balances cost and risk factors.
Solution Approach 2:
The patent transforms the path planning problem into a mathematical optimization problem by defining a cost function with multiple parameters (cable length, terrain difficulty, risk factors). The FMM algorithm iteratively updates distance and direction parameters across the computational grid to find the minimum cost path, enabling precise quantitative optimization rather than subjective manual evaluation.
2Measurement precision
If the Fast Marching Method (FMM) is used for path optimization, then precise numerical solutions can be obtained, but the sequential nature of the algorithm impedes direct parallelization
Solution Approach 1:
The patent divides the computational domain into multiple sub-domains that can be processed in parallel. Each sub-domain is assigned to a separate computational thread, allowing simultaneous execution of the FMM algorithm across different regions of the grid. This segmentation maintains the precision of the original FMM while enabling parallel computation to reduce overall processing time.
Solution Approach 2:
The patent performs preliminary domain decomposition and initializes boundary conditions for all sub-domains before parallel execution. By pre-processing the computational grid and setting up initial states, the system minimizes communication overhead during parallel execution and ensures that each thread can independently perform precise FMM calculations without waiting for sequential coordination.
3Device complexity
If sequential FMM is used for path planning, then algorithm simplicity is maintained, but computation time is excessive for ultra-long-distance routes
Solution Approach 1:
The patent segments the computational domain into multiple sub-domains that can be processed simultaneously in parallel threads. This segmentation increases algorithm complexity slightly by requiring domain decomposition and boundary condition management, but it dramatically reduces computation time for ultra-long-distance submarine cable routes by utilizing multi-core processors effectively.
Data Source
AI summary
A computer-implemented method for ultra-long distance path planning includes (a) decomposing a target area into two initial sub-domains, the target area covering a path from a starting point to an end point, at an initial resolution along a direction connecting the starting point to the end point, (b) executing a Fast Marching Method (FMM) in parallel in the two initial sub-domains to obtain an initial path connecting the starting point to the end point, (c) decomposing an area enclosing the initial path into a plurality of subsequent sub-domains at a resolution higher than the initial resolution, wherein the area enclosing the initial path is decomposed aligned with the direction connecting the starting point to the end point, and (d) executing the FMM in parallel in all the sub-domains to obtain an optimal path connecting the starting point and the end point.