In real-world industrial applications, bearings are typically operated under variable speeds and loads depending on the production condition, which results in nonstationary vibration signals from the bearings. Synchrosqueezing transform is a method that can effectively reflect the change in frequency with time, which is suitable for processing nonstationary bearing signals. However, significant classification features are difficult to extract from time–frequency information when operation conditions such as speed and load change frequently. Hence, an improved two-dimensional (2D) convolutional neural network (CNN) named the 2D multiscale cascade CNN (2D MC-CNN) is proposed for performing bearing fault diagnosis under various operating conditions. In a 2D MC-CNN, a multiscale information fusion layer is added prior to the convolutional layer of a conventional CNN to form MC images such that sensitive bands can be acquired for fault recognition. Experiments are conducted on bearings by considering various sets of fault categories and fault severity levels under six operating conditions. The experimental results show that the proposed method effectively extracts fault-related features and demonstrates excellent diagnostic accuracy and robustness. Comparisons with the original CNN and other typically used fault diagnosis methods based on the same dataset demonstrate the effectiveness of the proposed 2D MC-CNN and bearing fault diagnosis method.

To improve the vibration isolation performance for a parallel electromagnetic isolation system, an improved genetic algorithm to optimize the Q and R matrices in the control objective function for a model predictive control approach is proposed. In this study, a parallel electromagnetic isolation system with two electromagnetic isolation units is designed to expand the vibration isolation range to isolate the large object. The dynamical equation and state equation of the parallel electromagnetic isolation system are built. The nonlinear relationship among electromagnetic force, coil current, and gap is calculated by COMSOL Multiphysics to design the model predictive control controller. Meanwhile, an improved genetic algorithm by the variable chromosome length coevolutionary method is presented to tackle two issues. The first issue is that the parameters of Q and R matrices in the control objective function are mainly selected by trial and error. The other issue is that the model predictive control approach needs to determine prediction steps which may lead to the model predictive control approach suffering from heavy computation or an inaccurate prediction model. Simulation and experimental results demonstrate that the parallel electromagnetic isolation system with model predictive control method based on the improved genetic algorithm can achieve better vibration isolation performance in comparison with the passive isolation system.

Optimization scheduling plays a pivotal role in construction projects, significantly influencing both the overall project schedule and its efficiency. This study focuses on optimizing the scheduling of electric highway engineering projects within roadbed construction. The research considers multiple earthmoving processes and optimizes the working time of each piece of equipment, taking into account its capacity and speed within a limited working week. The study is further contextualized by the use of regional time‐of‐use (TOU) electricity pricing. A sophisticated optimization model is developed to simulate optimal machinery operation, striking a balance between energy consumption and work efficiency. This paper introduces an innovative algorithm, the improved crested porcupine optimizer (ICPO), which incorporates Latin hypercube sampling for population initialization. To enhance algorithmic effectiveness, a combined strategy of parallel and compact processing is employed. This approach reduces the number of iterations required for optimization and consequently lowers energy consumption. Rigorous analysis and comparison with existing algorithms demonstrate that ICPO significantly reduces both iteration count and financial expenditure. Simulation results validate the accuracy and practicality of the proposed model and algorithm, showing a reduction of over 7% in both engineering time and energy consumption.

In this study, the optimization of construction machinery scheduling within roadbed construction projects is explored, taking into account both personnel fatigue and sequence-dependent setup times. A sophisticated optimization model has been developed to simulate the optimal operation of machinery, aiming to maximize equipment utilization efficiency while addressing the challenges posed by worker fatigue. An innovative algorithm, the improved hybrid gray wolf and whale algorithm fused with a penalty function for construction machinery optimization (IHWGWO), is introduced, incorporating a penalty function to handle constraints effectively. This algorithm reduces the number of iterations required for optimization and, subsequently, cuts down on energy consumption. Through rigorous analysis and comparison with existing algorithms, the proposed IHWGWO demonstrates a significant reduction in both iteration count and financial expenditure. Simulation outcomes confirm the accuracy and practicality of the model and algorithm, establishing a promising new approach for scheduling in construction engineering.