This comprehensive study develops advantageous optimization methods to solve a nascent problem, namely multi-task simultaneous supervision dual resource-constrained (MTSSDRC) scheduling. MTSSDRC is a complex problem that deals with machine assignment, job sequencing, operator allocation, and task sequencing. Setup and unloading must be scheduled to operators, and they are allowed to leave machines while processing jobs. Earlier research on MTSSDRC developed a permutation-based genetic algorithm (PGA) with a specific decoding scheme, namely DSE, to solve the problem. Many previous studies succeed in solving scheduling problems by modifying well-known metaheuristic techniques. Therefore, we are inspired by this to explore further modifications to particular metaheuristics. The first contribution of the present study lies in the development of new decoding schemes that can perform better than the existing option. Five new decoding schemes are considered. Two of those schemes, namely DS2 and DS4, perform significantly better than DSE, reaching 6% relative deviation. DS4 is superior in terms of solution quality, but DS2 can run eight times faster. Another contribution is the development of six modified metaheuristics that are implemented for the MTSSDRC problem: tabu search, simulated annealing, particle swarm optimization, bees algorithm (BA), artificial bee colony, and grey wolf optimization. The performance of these metaheuristics is compared with that of the PGA. The results show that the PGA and BA are consistently superior for medium- and large-sized problems. The BA is more promising in terms of solution quality, but the PGA is faster.

这项全面的研究提出了有利的优化方法来解决一个新兴问题，即多任务同时监管双重资源受限（MTSSDRC）调度。MTSSDRC是一个复杂的问题，涉及机器分配，作业排序，操作员分配和任务排序。安装和卸载必须安排给操作员，允许他们在处理作业时离开机器。MTSSDRC的早期研究开发了一种基于排列的遗传算法（PGA），该算法具有特定的解码方案DSE，以解决该问题。许多以前的研究通过修改众所周知的元启发式技术成功解决了调度问题。因此，我们受此启发而探索了对特定元启发式方法的进一步修改。本研究的第一个贡献在于开发了新的解码方案，这些方案的性能要优于现有的方案。考虑了五个新的解码方案。其中两个方案DS2和DS4的性能明显优于DSE，相对偏差达到6％。DS4的解决方案质量优越，但DS2的运行速度快八倍。另一个贡献是开发了针对MTSSDRC问题实施的六种修改的元启发式算法：禁忌搜索，模拟退火，粒子群优化，蜜蜂算法（BA），人工蜂群和灰太狼优化。将这些元启发式方法的性能与PGA的性能进行比较。结果表明，PGA和BA在中型和大型问题上始终具有优越性。