With the significant growth of multiprocessor systems (MPS) to deal with complex tasks and speed up their execution, the energy generated as a result of this growth becomes one of the significant limits to that growth. Although several traditional techniques are available to deal with this challenge, they don’t deal with this problem as multi-objective to optimize both energy and makespan metrics at the same time, in addition to expensive cost and memory usage. Therefore, this paper proposes a multi-objective approach to tackle the task scheduling for MPS based on the modified sine-cosine algorithm (MSCA) to optimize the makespan and energy using the Pareto dominance strategy; this version is abbreviated as energy-aware multi-objective MSCA (EA-M2SCA). The classical SCA is modified based on dividing the optimization process into three phases. The first phase explores the search space as much as possible at the start of the optimization process, the second phase searches around a solution selected randomly from the population to avoid becoming trapped into local minima within the optimization process, and the last searches around the best-so-far solution to accelerate the convergence. To further improve the performance of EA-M2SCA, it was hybridized with the polynomial mutation mechanism in two effective manners to accelerate the convergence toward the best-so-far solution with preserving the diversity of the solutions; this hybrid version is abbreviated as EA-MHSCA. Finally, the proposed algorithms were compared with a number of well-established multi-objective algorithms: EA-MHSCA is shown to be superior in most test cases.

随着处理复杂任务并加快执行速度的多处理器系统（MPS）的显着增长，这种增长所产生的能量已成为该增长的重要限制之一。尽管有几种传统技术可以解决这一难题，但它们并没有将其作为多目标同时解决能源和制造跨度指标的多目标问题来解决，除了昂贵的成本和内存使用量之外。因此，本文提出了一种基于改进的正弦余弦算法（MSCA）的多目标方法来解决MPS的任务调度问题，并使用帕累托优势策略优化了制造时间和能量。此版本缩写为能量敏感型多目标MSCA（EA-M2SCA）。基于将优化过程分为三个阶段，对经典SCA进行了修改。第一阶段在优化过程开始时尽可能地探索搜索空间，第二阶段围绕从总体中随机选择的解决方案进行搜索，以避免陷入优化过程中的局部极小值，最后一步围绕最佳选择进行搜索。 -so-so解决方案可加速收敛。为了进一步提高EA-M2SCA的性能，它以两种有效的方式与多项式突变机制进行了杂交，以在保持解决方案多样性的同时，加速向最佳解决方案的收敛。该混合版本缩写为EA-MHSCA。最后，将提出的算法与许多公认的多目标算法进行了比较：在大多数测试案例中，EA-MHSCA被证明是优越的。