Blast furnaces play a critical role in the steel industry, and their operational optimization is crucial for energy conservation and emissions reduction. This paper examines the impact of changes in operational conditions on blast furnace performance. We propose a dynamic multi-objective optimization algorithm based on multiple short time series (MT-DC-RVEA) to solve the constructed dynamic multi-objective operational optimization model for blast furnaces. Experimental results validate the effectiveness of the proposed algorithm in solving the operational optimization model for blast furnace operations.
Under the condition of informatized joint combat system, the traditional target value analysis method based on human judgment is no longer able to quickly and accurately judge the target value from the massive battlefield information. In this paper, combining with the attention mechanism model, we propose a dual-driven target value assessment framework based on target and data, and construct data-driven and purpose-driven attention mechanism models respectively. By selecting multiple features and calculating the saliency of cue targets, the results of the two models are fused to screen out the noticed cue targets and further expanded to form a high-value target set. The experimental results show that the proposed target value assessment model based on attention mechanism is effective, and the matching rate between the actual strike targets and the high-value targets obtained by the model reaches 92%, which is in line with the actual situation of combat and can reflect the cognitive process of the commander to some extent.
Membrane computing provides efficient computing devices for broad applications due to its distributed storage and the parallel processing. As computing devices in membrane computing, numerical P systems with thresholds (NPT systems) are proven to be Turing universal, and their computational and operational semantics need to be further investigated. In this work, the intrinsic relationship between NPT systems and Petri nets is concerned. The ingredients of Petri nets are associated with the elements of numerical variables in NPT systems, and the operations of Petri nets are associated with the evolutions of NPT systems. The results on the boundedness and reachability of NPT systems are obtained by using the relationship between NPT systems and Petri nets.
For the latest two years, relation classification-based surrogate-assisted algorithms show good potential for solving expensive multi-objective optimization problems (EMOPs). In this category of methods, the used dominance relation that is vital for building training dataset and selecting promising solutions to reduce expensive real function evaluations (FEs). However, the existing studies are still at the initial stage and lack specific research on the dominance relation. This paper proposes a novel dominance relation called Difference Vector Angle Dominance with an angle threshold for EMOPs (called as DVAD- φ ). The proposed DVAD- φ has adaptive selection pressure and considers the convergence and diversity of solutions when picking out superior solutions, which makes it beneficial to pick out promising solutions for expensive real FEs and reduce expensive real FEs. To be specific, we firstly give the definition of DVAD- φ that measures the superiority from one solution to another solution, where the angle threshold φ controls the selection pressure. Then, we propose an adaptive determination strategy of angle threshold based on bisection to set proper pressure for picking out promising solutions for expensive real FEs. Experiments have been conducted on 7 test functions from one benchmark set. The experimental results have verified the effectiveness of DVAD- φ .
Global search is a fundamental task in optimization, aiming to find the optimal solution across the entire search space. To address the challenges in global search, a hierarchical competitive differential evolution algorithm is proposed. It uniquely incorporates a hierarchical competition mechanism and an adaptive differential mutation strategy based on competition outcomes, substantially enhancing global search. The proposed algorithm benchmarks on a total of 30 international test functions of CEC 2017 benchmark functions. The convergence accuracy, coupled with the outcomes of two nonparametric statistical tests, the Friedman test and the Wilcoxon signed-rank test, clearly demonstrates that HCDE exhibits competitive performance when compared to the other 14 efficient optimizers.
Food is closely related to national economy and people's livelihood. Rice is the largest grain crop in China, it is crucial to predict the loss rate of rice during processing to reduce food waste and ensure food security. This study first obtained the loss rate of rice processing through the recovery survey form of enterprises. Then, prediction was carried out using two common models: the BP neural network and multiple linear regression. Finally, the genetic algorithm was applied to optimize the BP neural network for further prediction and com-pared with the original models. The experimental results showed that the GA-BP model had higher prediction accuracy and smaller error compared to the first two models. It is valuable in reducing processing losses and maintaining food security.
Underwater gliders have become one of the iconic technologies in ocean environmental monitoring, which relies on adjusting buoyancy to achieve buoyancy control and gliding through the water using hydrodynamics. They can be used for long-term and wide-range observation and detecting complex ocean environments. Meanwhile, underwater gliders have the advantages of low cost, long endurance, reusability, and certainly trajectory control capabilities. However, in the current cooperative detection, due to the complexity and large disturbances of ocean currents, there exists a problem of high energy consumption and a non-optimal path in the underwater glider cluster. To address this issue, this paper combines cluster size optimization and path planning, proposes a method that combines the classic path optimization problem in the plane with the special motion mode of the glider, and then extends it to the optimization method in three-dimensional motion, and conducts relevant simulation experiments for verification. The simulation results show that the proposed method can effectively obtain the path with the minimum energy consumption of the glider under a three-dimensional trajectory.
Flexible job shop scheduling (FJSP) is crucial for automated production, ensuring efficiency and flexibility. In recent years, deep reinforcement learning (DRL) has achieved success in solving sequence decision-making problems. However, the efficiency of the generated scheduling plans is often constrained by the dependence of most DRL algorithms on priority dispatching rules (PDR). In order to enable the agent trained in DRL to autonomously choose operations and machines, this paper proposes an expert-guided deep reinforcement learning framework (EGDRL). Based on the representation of scheduling states using a disjunctive graph and an operation-machine topology graph, a graph neural network (GNN) is used to capture the complex relationships between operations and machines. More importantly, in the early stages of training, this paper introduces expert-guided solutions by using PDR to guide the action selection of reinforcement learning, which greatly improves the quality of decision-making. Experimental results consistently show that the proposed method outperforms traditional PDRs and other DRL algorithm. Notably, this superiority is observed across various scales, including larger instances. Additionally, the method exhibits robust performance on instances not encountered during training.
In order to discover the real demand of cruise ship passengers and find the connection between their demand and cruise ship space design, we conduct a questionnaire survey of the passengers' characteristics and their eating behavior. Through the analysis on the passengers' eating behavior by using K-means clustering algorithm on the basis of the survey data, different types of their behavior characteristics are recognized, and the connections between the passengers' characteristics and their eating behavior are discussed. The research results are helpful to understand cruise ship passengers' eating demand and improve the design concept of cruise ship.
At present, most of the control systems of the full swing ship adopt the classical PI control or open loop transfer method, which has low control accuracy and slow response speed. In this paper, the idea of intelligent distributed control is introduced into the control of the propulsion device of the full swing ship pod, and the physical model and mathematical model of the hydraulic propulsion system of the pod are established. The PID control strategy based on fuzzy reasoning is designed for the full rotary propulsion device, which can accurately and reliably control the rotation Angle of the pod. Finally, the modeling and simulation of the full rotary hydraulic propulsion system of the pod are carried out under small Angle conditions. The results show that the rotation Angle error is small, and the proposed method has application value.
Multi-Objective UAV path planning problem is to find an optimal path, which can satisfy multiple objectives at the same time and optimize other performance indicators in the case of considering multiple conflicting objectives, constraints and trade-offs. In order to solve the challenge of considering multiple constraints and responding to environmental changes in real time, we propose a UAV path planning method based on multi-objective dwarf mongoose optimization algorithm. Considering the general search efficiency and insufficient global convergence of the original algorithms, nonlinear factorial augmented search strategies, chaotic mapping, and on-the-fly search strategies have been proposed to address the above problems and to enhance the algorithmic population diversity and local optimization capability. Experimental results show that, compared with other algorithms, the proposed method has better performance and robustness in multi-target UAV path planning, and can effectively find high-quality non-inferior solution sets, which provides an effective solution for UAV path planning.
It is difficult to control the rudder Angle of UUV due to its nonlinear characteristics, dynamic matching, and many interference terms. With the increase of the number and type of uncertainty items, the corresponding system interference of traditional PID controller increases sharply, which limits its application in engineering. In this paper, the rudder Angle control method based on sliding mode control is adopted. By establishing the UUV mathematical model and combining with the backstepping method, the control efficiency can be improved while having strong robustness. The simulation results verify the effectiveness of the proposed control method.
Genetic algorithms (GA) are widely used to solve complex combinatorial optimization problems, such as the Traveling Salesman Problem (TSP). However, as the problem being scale-up, the feasible solution space expands extremely large, resulting in a significant decrease in the efficiency of GA. This paper is dedicated to address this challenge on large TSP by proposing an improved Hierarchical Genetic Algorithm with Density-Based Clustering (HGADC). First the Hierarchical Density-Based Clustering (HDBSCAN) algorithm was employed to cluster all cities into clusters. Each cluster will be considered as a sub-problem, by this way we have a hierarchical structure, i.e. low-level intra-cluster sub-problems and high-level inter-clusters problems. For low-level, a novel Elite Ranking strategy GA was proposed. While for high-level, we designed two approaches, one is the Barycenter Method (BM), which condensed the cluster to be an individual city, the other is called Gene Segment Method (GSM), which kept both head and tail of gene segment as a representation of the cluster. Finally, the results of two levels were integrated to form the final solution. HGADC facilitated the preservation of high-quality genes in GA via the clustering method combined with GA, dramatically reduced the problem's size and computational costs. The experiments showed that the HGADC algorithm exhibits promising performance in solving large TSPs in TSPLIB compared with notable methods such as MMAS, PACO-3OPT, and KIG.
With the rapid development of information technology making the scale of the Internet increasing day by day, collaborative optimization of multiple scheduling tasks in a multi-cloud environment provides users with faster scheduling options. Meanwhile, there is a certain similarity between cloud scheduling tasks, and in order not to waste the similarity between tasks, similar tasks are linked together to find an optimal scheduling solution for multiple tasks, making it possible to handle multiple scheduling tasks simultaneously. Firstly, we construct a multi-objective optimization model considering time, cost and VM resource load balance; secondly, since there are not only independent optimization problems in real scenarios, we adapt the constructed multiple similar optimization models and propose a multi-task multi-objective optimization model; finally, to be able to solve the constructed model better, we use a proposed objective function-based Finally, we propose an evolutionary multitasking algorithm based on weighted summation of the objective functions, which allows the algorithm to find the optimal solution among multiple multi-objective models. Simulation experiments show that the proposed algorithm has better performance.
To reduce the impact of greenhouse effect, the deployment and utilization of renewable energy sources such as wind power has become an inevitable trend. Therefore, the decision of economic emission dispatch (EED) problems is particularly important. In this paper, to incentivize renewable energy generation, the green certificate trading mechanism is introduced to solve from the economic level. However, as the dimensions of the EED problems are increased, the current methods cannot make proper scheduling decisions in a short time. Therefore, the EED problems are categorized as computationally expensive EED problems. To solve the above problems, a surrogate-based multi-objective optimization method is proposed. On one hand, the artificial neural network (ANN) surrogate models are proposed to replace the traditional objective function, which greatly reduces the time to obtain feasible decisions. On the other hand, a modified multi-objective gray wolf optimizer (MOGWO) is proposed to execute EED optimization accurately and quickly. This algorithm improves the search ability and convergence of the original MOGWO algorithm through improving the position update strategy and introducing the difference algorithm. The effectiveness of the surrogate-based MOGWO is testified through simulations of benchmark functions and computationally expensive EED optimization problems within the actual Taipower 40-unit test system.
With the rapid development of DNA nanotechnology, a variety of flexible molecular computing models and logic computing systems have been proposed and constructed. Among them, the switching circuit model based on DNA strand displacement technology performs well in terms of the decreased complexity in constructing multi-input and multi-output circuits. However, in the previous established circuits based on the switching computing model, the same switching elements are designed differently to generate diverse target strands for different outputs, increasing the complexity and difficulty of the design. Here, a top-down design strategy is proposed in which only once design is required for each switching element. The presented strategy decreased the amount of switching elements and simplifies the design complexity to a certain extent. To verify the strategy, the 4-2 priority and 10-4 priority encoder is realized by this design and validated by Visual DSD. The proposed method has possibilities in design of large-scale DNA logic computing systems with more inputs and outputs.
In this paper, a reinforcement learning-based differential evolution algorithm with levy flight strategy (RLLDE) for solving optimization problems is proposed. It introduces a novel mutation mode considering search directions is proposed firstly. Secondly, a levy flight strategy is employed to enhance the exploration capability of Differential Evolution (DE). Lastly, the Q-learning method from reinforcement learning is introduced to establish a switching mechanism between two different updating modes during the mutation stage. These strategies effectively improve the algorithm's convergence speed and accuracy. RLLDE is analyzed on CEC 2017 benchmark functions to validate its optimization performance. Compared to five basic DE and eight efficient optimizers, the experimental results demonstrate that the algorithm exhibits efficient and effective performance in solving optimization problems.
Solving multi-objective optimization problems with constraints is a complex task, involving the delicate balance of interrelated and conflicting objective functions and constraint values. While numerous evolutionary algorithms have been developed to tackle this issue, they often struggle to effectively balance population feasibility, convergence, and diversity. In this paper, we introduce the bi-population dynamic coordinated non-dominated sorting algorithm (DDCNDS), which maintains two dynamically coordinated populations and can adjust population size at different evolution stages to balance feasibility and convergence. During the early stages of evolution, the focus is primarily on convergence, with the first population having a larger size to facilitate convergence to the feasible domain across infeasible regions. In the later stages, the emphasis shifts to feasibility, with the second population having a larger size and utilizing local search to explore the feasible domain. Additionally, we propose a non-dominated criterion sorting method to select better individuals, adjusting the non-dominated level based on the proportion of feasible solutions in the input population. Experimental results on three well-known benchmark suites demonstrate the competitiveness of the proposed algorithm.
Particle swarm optimization (PSO) is a class of modern generalized intelligent optimization algorithms. Comprehensive learning PSO (CLPSO) is a powerful PSO variant that is good at exploration and works well on many multimodal problems. Enhanced CLPSO (ECLPSO) and adaptive CLPSO (ACLPSO) are improved versions of CLPSO that we have previously proposed. ECLPSO enhances the exploitation performance of CLPSO, and ACLPSO strengthens the exploration performance of ECLPSO. We have compared CLPSO, ECLPSO and ACLPSO on 16 benchmark functions in our previous study. To further understand the generalization performance of the three algorithms, this paper compares them on the well-known CEC2013 test set of multimodal benchmark functions. Compared with the benchmark functions employed in the previous study, though the number of dimensions for the CEC2013 benchmark functions are quite smaller, there are also a number of local optima in the search space. In addition, the CEC2013 test set contains composition functions that mix different characteristics of various basic functions, causing the search space to have a huge quantity of local optima and is very complex. Experimental results demonstrate that the accuracy of the solution obtained by ECLPSO is slightly better than that by CLPSO on just a few functions and is similar on all the other functions; both ECLPSO and CLPSO fail to derive the global optimum or a near-optimum on most of the composition functions; and ACLPSO is well balanced between exploration and exploitation, as it outperforms ECLPSO and CLPSO by being able to find the global optimum or a near-optimum with high accuracy on almost all the functions.
Speed and Resistance are very critical general parameters for ship hull forms design. Reducing resistance is one of the main methods for obtaining higher speed. Optimal hull form contributes a lot to the resistance and wave-making of a vessel. In recent research, the effectiveness of resistance reducing by twin-bow appendage for monohull ship has been paid high attention. The twin-bow appendage has been developed for high-speed monohull vessels to reduce the resistance. However, for some monohull ships such as monohull with bulbous bow it is impossible to deploy twin-bow appendage since it is necessary to consider the load capacity of bulbous bow. For exploring the applicability of twin-bow appendage, a twin-bow appendage with drum is proposed for replacement of bulbous bow in this article. The hydrostatic resistance of normal vessel with bulbous bow and that of twin-bow appendage with drum will be analyzed and compared by numerical simulation. RANS equation is applied for numerical computation and the CFD tool Star CCM+ is used for simulation.
