Pointing Scheduler Grouping and Dynamic Programming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in generating efficient pointing schedules for pointable devices within acceptable computation times, considering the relative directions, dwell times, weights, windows of performance, and other time constraints for multiple targets.
Innovation Solution
The method involves partitioning targets into groups based on angular and physical distances, overlapping windows of performance, and similar weights, then determining group schedules and using a grow and prune approach to create a complete pointing schedule.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If an optimized pointing schedule is generated to minimize slew times and maximize target inclusion, then the productivity and efficiency of the pointable device is improved, but the computation time and complexity of the scheduling algorithm increases
Solution Approach 1:
The patent divides the target queue into multiple groups based on spatial proximity and temporal overlap of performance windows. Each group is processed independently through dynamic programming, breaking the overall complex scheduling problem into smaller subproblems that can be solved more efficiently and then combined to form the complete optimized schedule.
Solution Approach 2:
The patent performs preliminary clustering of targets into groups based on spatial and temporal criteria before executing the dynamic programming optimization. This pre-processing step organizes the data structure to facilitate more efficient computation during the subsequent optimization phase, reducing the overall computational burden.
2Reliability
If all targets in the target queue are included in the pointing schedule, then the completeness of observations is improved, but the total slew time and computation resources increase
Solution Approach 1:
The patent applies dynamic programming to find the optimal subset of targets to observe within the given time constraints and performance windows. Rather than forcing inclusion of all targets, the algorithm determines the maximum value subset that can be realistically achieved, allowing partial completion when full coverage is not feasible while still maximizing overall effectiveness.
Solution Approach 2:
The patent transforms the scheduling problem by changing the parameter representation from individual target sequences to group-based state transitions. By defining states based on group completions and using value functions to represent cumulative effectiveness, the algorithm simplifies the complexity of evaluating all possible target permutations while maintaining optimality.
3Productivity
If the pointing device redirects frequently between multiple targets, then the target coverage is improved, but the slew time consumption increases
Solution Approach 1:
The patent merges adjacent targets in the target queue into groups when their performance windows overlap and they are spatially proximate. By treating multiple targets as a single group unit, the device can observe them in sequence without returning to the home position between targets, thereby reducing total slew time while maintaining comprehensive coverage of all grouped targets.
Solution Approach 2:
The patent introduces a group-based dimensional organization to the scheduling problem, where targets are arranged not just by their sequential order but by spatial proximity and temporal overlap. This additional organizational dimension allows the algorithm to optimize the observation sequence by grouping targets that can be efficiently visited together, reducing redundant movements.
Data Source
AI summary
A method of determining a schedule for a pointing device includes distributing targets among a plurality of groups and assigning schedules to the groups, after which the groups and group schedules are not modified. A partial pointing schedule is successively grown by selecting a list of remaining groups, and appending a best candidate sub-schedule for the list to the partial schedule. If a sub-schedule cannot be found for which the grown partial schedule meets all applicable requirements, the partial schedule is pruned to a previous state and a different, previously evaluated sub-schedule is appended to the pruned partial schedule to attempt re-growth. Groups, group schedules, lists, and sub-schedules can be selected according to values, dwell times, and/or windows of performance of the targets, groups, and sub-schedules. If no schedule is found that meets all requirements, a missed group can be excluded from the schedule.