Optimization of Agent Patrol Algorithm for Patrolling Set of Points with Different Frequency

The design of agent patrol system with different task targets needs to develop reasonable patrol algorithms and strategies. This work proposes an efficient algorithm to optimize robotic patrol system based on the pinwheel scheduling algorithm and the greedy algorithm with optimal characteristics. It can perform patrolling tasks according to the checkpoints number and the visit frequency provided by the user. One-dimensional and two-dimensional scheduling algorithms are designed for patrolling checkpoints in linear and planar distributions respectively. To assess the effectiveness of the optimized algorithms, a comparison test with traditional pinwheel algorithm is conducted, which shows that the time complexity of the combined algorithm is reduced from O(n²) to O(log n), significantly improving the patrolling efficiency. In real-world test, the combined algorithm reduces the running time by about 25%, while the peak memory usage is reduced by nearly 17%, proving that the algorithm has better stability and scalability in resource-constrained systems.