The artificial potential field approach is an efficient path planning method. However, to deal with the local-stable-point problem in complex environments, it needs to modify the potential field and increases the complexity of the algorithm. This study combines improved black-hole potential field and reinforcement learning to solve the problems which are scenarios of local-stable-points. The black-hole potential field is used as the environment in a reinforcement learning algorithm. Agents automatically adapt to the environment and learn how to utilize basic environmental information to find targets. Moreover, trained agents adopt variable environments with the curriculum learning method. Meanwhile, the visualization of the avoidance process demonstrates how agents avoid obstacles and reach the target. Our method is evaluated under static and dynamic experiments. The results show that agents automatically learn how to jump out of local stability points without prior knowledge.

Industrial products' reuse, recovery, and recycling are very important because of their environmental and economic benefits. Effective product disassembly planning methods can improve their recovery efficiency and reduce their bad environmental impact. However, the existing approaches pay little attention to sequence-dependent disassembly with resource constraints, such as limited disassembly operators and tools, which makes the current planning methods ineffective in practice. This paper considers a multiobjective resource-constrained and sequence-dependent disassembly optimization problem with disassembly precedence constraints. Energy consumption is adopted to evaluate the disassembly efficiency. Its use with traditional optimization criterion leads to a novel multiobjective optimization model such that the energy consumption and disassembly time are minimized while disassembly profit is maximized. Since the problem complexity increases with the number of components in a product, a lexicographic multiobjective scatter search (SS) method is proposed to solve the proposed multiobjective optimization problem. Its effectiveness is verified by comparing the results of linear weight SS and genetic algorithms. The results show that it is able to provide a better solution in a short execution time and fulfills the precedence requirement in a product structure and resource constraints.

Recycling, reusing, and remanufacturing of end-of-life (EOL) products have been receiving increasing attention. They effectively preserve the ecological environment and promote the development of economy. Disassembly sequencing and line balancing problems are indispensable to recycling and remanufacturing EOL products. A set of subassemblies can be obtained by disassembling an EOL product. In practice, there are many different types of EOL products that can be disassembled on a disassembly line, and a high-level uncertainty exists in the disassembly process of those EOL products. Hence, this paper proposes a stochastic multi-product multi-objective disassembly-sequencing-line-balancing problem aiming at maximizing disassembly profit and minimizing energy consumption and carbon emission. A simulated annealing and multi-objective discrete grey wolf optimizer with a stochastic simulation approach is proposed. Furthermore, real cases are used to examine the efficiency and feasibility of the proposed algorithm. Comparisons with multi-objective discrete grey wolf optimization, non-dominated sorting genetic algorithm II, Multi-population multi-objective evolutionary algorithm, and multi-objective evolutionary algorithm demonstrate the superiority of the proposed approach. Note to Practitioners —Disassembly line balancing has been widely recognized as the most ecological way of retrieving EOL products. Through in-depth research, we present a Stochastic Multi-product Multi-objective Disassembly-sequencing-line-balancing Problem. Furthermore, we consider that the uncertainty of products might cause disassembly failure. To solve this problem effectively and quickly, we combine the simulated annealing algorithm with the Grey Wolf Optimizer. The results show that the algorithm can effectively solve the proposed problem. The disassembly scheme provided by the obtained solution set offers a variety of options for decision-makers.

Industrial products' reuse, recovery, and recycling are very important due to the exhaustion of ecological resources. Effective product disassembly planning methods can improve the recovery efficiency and reduce harmful impact on the environment. However, the existing approaches pay little attention to disassembly resources, such as tools and operators that can significantly influence the optimal disassembly sequences. This article considers a multiobjective resource-constrained disassembly optimization problem modeled with timed Petri nets such that energy consumption is minimized, while disassembly profit is maximized. Since its solution complexity has exponential growth with the number of components in a product, a multiobjective genetic algorithm based on an external archive is used to solve it. Its effectiveness is verified by comparing it with nondominated sorting genetic algorithm II and a collaborative resource allocation strategy for a multiobjective evolutionary algorithm based on decomposition.

Production scheduling is a crucial task in modern steel plants. The scheduling of a wire rod and bar rolling process is challenging in many steel plants, which has a direct impact on their production efficiency and profit. This article studies a new single-machine scheduling problem with sequence-dependent setup time, release time, and due time constraints originated from a wire rod and bar rolling process in steel plants. In this problem, jobs have been assigned to batches in advance. The objective is to schedule the batches and jobs on continuous time to minimize the number of late jobs. A mixed-integer program is created as a baseline model. A baseline method is used to solve this NP-hard problem by solving the baseline model. We further design a two-stage decomposition method after analyzing the characteristics of this problem. Both actual and simulated instances with varying sizes are solved by using the proposed methods. The results demonstrate that the baseline method can only solve some small-scale cases, while the decomposition method can solve all small-scale cases and some medium-scale cases. Finally, we reveal the impacts of different instances on the performance of the proposed decomposition method.