With the continuous improvement of the intelligence level of industrial systems, equipment fault diagnosis in dynamic environments is confronted with issues such as the emergence of new fault modes and data distribution shift. Traditional batch training models struggle to adapt to various persistent changes. Incremental learning can integrate new data without forgetting acquired knowledge, offering a groundbreaking solution for the field of fault diagnosis. This paper systematically reviews the research progress of incremental learning in fault diagnosis. First, it introduces the basic theories of incremental learning and categorizes it into class-incremental, domain-incremental, task-incremental, and data-incremental learning. Second, it analyzes the practical applications of regularization, dynamic architecture, memory replay, and meta-learning techniques in incremental learning scenarios. Finally, it prospects future research directions including open-set incremental diagnosis and interpretable incremental learning. This review intends to serve as a reference for researchers in this field and advance the development of incremental fault diagnosis.
Efficient execution of scientific workflows on cloud infrastructures requires scheduling decisions that can handle several conflicting requirements at the same time, including completion time, execution cost, energy use, and resource utilization. This study introduces a neuromorphic scheduling approach, named the Pareto Spiking-Hopfield Scheduler (PSHS), in which task-to-virtual-machine allocation is guided by spiking dynamics and constraint-aware control mechanisms. Instead of relying on conventional population operators, the proposed scheduler searches the assignment space through spiking neuron activity, while feasibility is supported by dedicated mechanisms for unique task assignment, capacity control, and budget monitoring. Different preference settings are explored to obtain a set of non-dominated scheduling alternatives. The approach is assessed using standard scientific workflow benchmarks, namely CyberShake, Montage, Epigenomics, and Inspiral, and is compared with NSGA-II, NSGA-III, and MOPSO under an equal optimization budget. The experimental results show that the behavior of PSHS depends on the workflow structure and instance size. The strongest improvements are observed on Epigenomics instances, where PSHS provides better Pareto-front quality and substantially lower IGD values. In particular, IGD is reduced by 98.6
Internet of Everything (IoE) services generate heterogeneous, bursty, delay-sensitive, and energy-intensive tasks that must be scheduled on virtualized cloud platforms under coupled delay, energy, SLA, migration, and utilization constraints. This paper presents ER-RL-DVFS, a hierarchical energy-risk-aware scheduling framework for DVFS-enabled IoE cloud systems. The framework combines an energy-risk task-priority score, DVFS-aware cluster–PM–VM profiling, hierarchical candidate filtering, deterministic score-based assignment for clearly separated candidates, selective tabular reinforcement learning as the prescribed refinement mechanism for near-tie decisions, and migration-aware correction. The methodological contribution is the coordinated scheduling pipeline rather than a new reinforcement-learning architecture. The numerical study uses a controlled stochastic benchmark with frozen seed-level outputs for five independent seeds (7, 19, 41, 73, 101), two resource configurations ( 512× 16 and 1024× 32 ), generic IoE and Smart-City IoT workloads, and eight baselines, including both GIoTDVFS-SFB and GIoTDVFS-mGA. Reported curves and heavy-load summaries use seed means and Student-t 95 512× 16 , ER-RL-DVFS obtains a mean makespan of 1068.47± 16.77 and energy consumption of (3.722± 0.035)× 10^5 J, compared with 1329.22± 12.37 and (4.305± 0.077)× 10^5 J for DQN-VMPlacement. The corresponding mean energy-delay product decreases from 0.5722× 10^9 to 0.3977× 10^9 , a 30.50
Wireless Mesh Networks (WMNs) have undergone significant development in recent years due to their low cost, ease of implementation, and dependable service coverage. However, the placement of nodes in these networks presents a crucial research problem for network operators and has a significant impact on the performance of WMNs. This issue is known to be NP-hard, and the use of approximate optimization algorithms, such as meta-heuristic techniques, is essential for its resolution with reasonable time execution. In this paper, three approaches were suggested to solve the Mesh Routers (MRs) placement problem with service priority. The first one is an adaptive version of Snake Optimizer (SO), called GSO, based on the incorporation of the Generalized Opposition Based-Learning (GOBL) strategy into the original SO to improve its exploration phase. The second one is a hybrid approach, called SOSA, based on the combination of SO with Simulated Annealing (SA). The last one is a hybrid approach, called ESOSA, based on the integration of GOBL mechanism and SA algorithm into the original SO, to improve both exploration and exploitation phases. A new problem formulation is suggested, where the main objective is to determine the optimal positions of the MRs in order to maximize the weighted coverage while satisfying the connectivity requirement, depending on a given set of Mesh Client (MCs) weights and locations. This is particularly relevant as most previous studies on MRs placement have treated all MCs equally, neglecting those with higher priority needs. The proposed approaches are validated in terms of fitness value while varying three key parameters: the number of MRs, the number of MCs, and the Transmission Range (TR) value. Simulation results obtained using MATLAB demonstrate that the proposed approaches outperform the standard SO and SA algorithms, as well as Particle Swarm Optimization (PSO), Grey Wolf Optimizer (GWO), Whale Optimization Algorithm (WOA), Firefly Algorithm (FA), and Harris Hawks Optimization (HHO). Furthermore, ESOSA achieves superior effectiveness in optimizing weighted coverage compared to all evaluated algorithms.
Energy-efficient task scheduling is a critical challenge in heterogeneous cloud data centers, where makespan, energy consumption, runtime, and resource feasibility must be optimized simultaneously. This paper proposes a constraint-aware hybrid metaheuristic framework for multi-objective cloud task scheduling by integrating Genetic Algorithm (GA), Grey Wolf Optimizer (GWO), threshold-based local refinement, and Colored Petri Net (CPN)-based event-driven simulation. The CPN module formally models task dispatching, parallel execution, CPU, RAM, and bandwidth contention, resource allocation, task completion, and waiting behavior, enabling feasibility-aware evaluation under concurrent execution. GA and GWO first explore the scheduling search space, and the best solutions are then refined through a threshold-guided local improvement mechanism without relying on penalty functions. Experimental results show that the proposed Hybrid method outperforms GA and GWO, achieving the lowest mean fitness of 674.20, makespan of 821.96, energy consumption of 452.56, and runtime of 10.20. It also records the smallest fitness standard deviation of 25.22, confirming higher stability. Statistical tests confirm significant improvements, with p-values of 1.024455e-07 against GA, 1.862645e-09 against GWO, and 2.066489e-10 for the Friedman test. Ablation, convergence, sensitivity, Pareto, ranking, energy, and scalability analyses further verify its robustness.
