模糊人工蜂群算法的旅行商问题求解

针对传统人工智能算法早熟收敛问题, 基于模糊化处理和蜂群寻优的特点, 提出一种模糊人工蜂群算法, 将模糊输入/输出机制引入到算法中来保持蜜源访问概率的动态更新。根据算法计算过程中的不同阶段对蜜源访问概率有效调整, 避免算法陷入局部极值。通过对旅行商问题的仿真实验和与其他算法的比较来验证算法的性能。计算结果表明, 该算法有良好的鲁棒性和有效性。     

求解旅行商问题的位置-次序编码差分演化算法

首先利用“差异算子”和“选择算子”描述了差分演化算法(DE)的基本原理,然后提出了一种新的、通用的特殊编码方法：位置 次序编码法,并利用此编码方法,提出了求解著名旅行商问题的离散差分演化算法：基于位置 次序编码的差分演化算法(PODE)。对于TSPLIB中两个不同规模的旅行商问题实例的计算表明,PODE算法具有极好的收敛性和稳定性     

A modified ant system to achieve better balance between intensification and diversification for the traveling salesman problem

This paper presents a new variant of Ant Colony Optimization (ACO) for the Traveling Salesman Problem (TSP). ACO has been successfully used in many combinatorial optimization problems. However, ACO has a problem in reaching the global optimal solutions for TSPs, and the algorithmic performance of ACO tends to deteriorate significantly as the problem size increases. In the proposed modification, adaptive tour construction and pheromone updating strategies are embedded into the conventional Ant System (AS), to achieve better balance between intensification and diversification in the search process. The performance of the proposed algorithm is tested on randomly generated data and well-known existing data. The computational results indicate the proposed modification is effective and efficient for the TSP and competitive with Ant Colony System (ACS), Max-Min Ant System (MMAS), and Artificial Bee Colony (ABC) Meta-Heuristic.     

求解旅行商问题的探索—开发—跳跃策略单亲遗传算法

针对遗传算法求解旅行商问题(TSP)时容易早熟、收敛速度慢等问题,提出一种基于探索—开发—跳跃策略的单亲遗传算法(EDJS-PGA)。该算法将基因移位、倒序、交换三种算子组合构成探索策略,用于扩展解的搜索空间,增强算法全局搜索能力;再将logistic混沌映射和改良圈操作融合为一种混沌映射改良圈算子,用于增强算法的局部搜索能力,构成开发策略;最后针对种群中的同优个体设计了近邻变异算子,构成跳跃策略,增强了算法跳出局部最优解的能力,使其兼具个体变异、局部优化、防止早熟等多重作用。通过对18个TSP实例进行仿真实验,结果表明EDJS-PGA相较于传统单亲遗传算法具有更高的求解精度和收敛速度,且最优解偏差率和平均误差率均处于较低水平;与其他文献对比,EDJS-PGA具有更强的鲁棒性和求解效率。     

竞争合作型协同进化免疫算法及其在旅行商问题中的应用

为提高人工免疫算法的收敛性能,提出了一种竞争合作型协同进化免疫优势克隆选择算法(CCCICA).把生态学中的协同进化思想引入到人工免疫算法中,考虑了环境和子群间相互竞争的关系,子种群内部通过局部最优免疫优势,克隆扩增,自适应动态高频混合变异等相关算子的操作加快了种群亲和度成熟速度.把信息熵理论引入到算法中完善了种群的多样性.所有子种群共享同一高层优良库,并将其作为抗体子种群领导集合,对高层优良种群进行免疫杂交操作,通过迁移操作把优良个体返回到各子种群,实现了整个种群信息交流与协作.针对旅行商问题(traveling salesman problem,TSP)多个实例结果表明:与其它智能算法相比较该算法具有较好的性能.     

A hybrid genetic algorithm for the traveling salesman problem with pickup and delivery

In this paper, a hybrid genetic algorithm （GA） is proposed for the traveling salesman problem （TSP） with pickup and delivery （TSPPD）. In our algorithm, a novel pheromone-based crossover operator is advanced that utilizes both local and global information to construct offspring. In addition, a local search procedure is integrated into the GA to accelerate convergence. The proposed GA has been tested on benchmark instances, and the computational results show that it gives better convergence than existing heuristics.     

基于混合蚁群算法的WTA问题求解

武器-目标分配问题(Weapon-TargetAssignmentProblem)是一种典型的NP问题。该文提出了一种基于遗传算法和蚁群算法的混合算法(GAACO)以解决武器-目标分配问题。首先,使用遗传算法对火力分配问题形成初始解;然后,将遗传算法的结果传递给改进的蚁群算法,对问题求精确解。实验结果表明该算法求精度优于遗传算法,时间性能优于传统蚁群算法。     

A new soft computing model based on linear assignment and linear programming technique for multidimensional analysis of preference with interval type-2 fuzzy sets

Multiple criteria decision making (MCDM) approach plays an important role in life, since it is always necessary to make decisions through various alternatives based on specific criteria. In this paper, interval type-2 fuzzy sets (IT2FSs) are used because in most cases in the real-world the information is incomplete and ambiguous. A new group decision approach with linear assignment method (LAM) is proposed. In addition, weight of each evaluation factor according to subjective and objective data is constructed based on a new developed version of linear programming technique for multidimensional analysis of preference (LINMAP) method. In the proposed method, weights of decision makers (DMs) are computed based on a novel approach that applies a new modified method based on the concept of ideal solutions. Furthermore, a new IT2F-ranking method is introduced. To display the applicability of the presented soft computing method, firstly, a real case study of green supplier selection problem is adopted from the literature and solved. Moreover, the method is applied in a second case study of project evaluation and selection problem. Two applications show that the introduced method presents a proper soft computing framework that can handle real-world uncertain environments. Moreover, the method can consider importance of the DMs and evaluation criteria.