基于膜计算与粒子群算法的盲源分离方法

为了解决盲源分离方法收敛速度慢、分离性能不高的问题,提出一种基于膜计算(Membrane Computing,MC)和粒子群算法(Particle Swarm Optimization,PSO)的盲源分离方法。算法以分离信号负熵作为粒子群的适应值函数,将粒子均匀分布到各基本膜中,将各基本膜内最优位置输出到表层膜并选择适应值最小的最优位置作为群体最优位置,通过粒子自身最优位置和群体最优位置对种群粒子进行速度和位置的更新。粒子群最优解调整盲源分离的步长函数,进行信号的分离。提出的算法简化了惯性权重取值问题,保证了PSO算法局部搜索的精度,满足了全局搜索的多样性。仿真实验和实例应用表明,提出的算法可以很好地分离混合信号,并且能避免PSO算法的早熟收敛问题,具有更快的收敛速度和更优异的分离性能。     

一种改进的雁群扩展粒子群算法研究

针对粒子群算法易陷入局部最优的问题,结合雁群启示粒子群算法和扩展粒子群算法提出了基于雁群启示的扩展粒子群(GeEPSO)算法。该算法在利用雁群飞行方向的多样性同时融合了所有粒子的个体极值信息,提高了种群多样性。为进一步提高改进算法的收敛速度,引入简化粒子群提出了 GeESPSO算法。基准函数的仿真表明:改进算法GeESPSO较好地平衡了收敛速度和局部最优两个矛盾,总体较优。为进一步验证算法在实际应用中的有效性,又分别用两种改进算法优化BP神经网络,并用相关气象数据对PM2.5的值进行预测。     

基于粒子群与聚类的多目标优化算法

针对粒子群优化算法容易陷入局部最优的问题,提出了一种基于粒子群优化与分解聚类方法相结合的多目标优化算法。算法基于参考向量分解的方法,通过聚类优选粒子策略来更新全局最优解。首先,通过每条均匀分布的参考向量对粒子进行聚类操作,来促进粒子的多样性。从每个聚类中选择一个具有最小聚合函数适应度值的粒子,以平衡收敛性和多样性。动态更新全局最优解和个体最优解,引导种群均匀分布在帕累托前沿附近。通过仿真实验,与4种粒子群多目标优化算法进行对比。实验结果表明,提出的算法在27个选定的基准测试问题中获得了20个反世代距离(IGD)最优值。     

基于信息熵改进的粒子群算法在桁架结构

提出了信息熵改进的粒子群优化算法用于解决有应力约束、位移约束的桁架结构杆件截面尺寸优化设计问题.首先介绍了信息熵基本理论和基本粒子群优化算法理论,然后对粒子群优化算法作了合理的参数设置,并将信息熵引入粒子群优化算法的适应函数和停机判别准则中.最后对2个经典的优化问题进行求解并与其他算法进行了比较.数据结果表明信息熵改进后的粒子群优化算法在桁架结构优化设计中优于其他同类算法.     

基于加速粒子群算法的车辆座椅悬架最优控制研究

针对传统最优线性二次型控制器中加权矩阵往往由设计者根据经验确定的问题,提出一种应用加速粒子群算法确定加权矩阵的方法。建立"车轮-车身-座椅、人体"6自由度随机振动系统模型,采用加速粒子群算法对座椅悬架进行参数优化,并对优化后系统进行最优线性二次型控制。将基于加速粒子群算法的最优线性二次型座椅悬架系统中"座椅、人体"垂向加速度与初始系统及基于常规粒子群算法和遗传算法的最优线性二次型控制系统进行对比,验证了此控制系统的有效性和优越性。     

基于混合粒子群算法的物流配送路径优化问题研究

针对物流配送路径优化问题,提出了一种融合Powell局部寻优算法和模拟退火算法的混合粒子群算法,以克服单用粒子群算法求解问题早熟收敛的不足,增加算法的开发能力,提高算法的全局搜索能力,并进行了实验计算.计算结果表明,用混合粒子群算法求解物流配送路径优化问题,可以在一定程度上提高粒子群算法在局部搜索能力和搜索全局最优解概率,从而得到质量较高的解.     

基于PSO-GRNN的直升机旋翼不平衡故障诊断

为了准确诊断直升机旋翼不平衡故障,提出了一种基于粒子群算法和广义回归神经网络模型(PSO-GRNN)的故障诊断方法。将交叉验证得到的平均均方误差作为粒子群的适应度函数,运用粒子群算法搜寻最优的GRNN光滑因子,建立最优的故障诊断模型。结果表明:采用PSO-GRNN模型可实现直升机旋翼不平衡的类型和程度的有效诊断,故障类型准确率高达94.29%,故障程度的诊断最大误差仅6.54%,满足工程需求。     

基于混沌粒子群的资源受限项目调度问题

鉴于基本粒子群算法易陷入局部最优,提出一种将混沌算法嵌入基本粒子群的混沌粒子群算法,并将其用于求解典型的资源受限项目调度问题。采用基于优先值的粒子编码方式和串行调度方案,分别用基本粒子群算法和混沌粒子群算法对实例求解。并且比较了2种算法求解多资源受限项目调度问题的性能。结果表明:混沌粒子群算法在距最优值的平均偏差和达到最优值的次数百分比等性能上要优于基本的粒子群算法,并且混沌粒子群具有更好的收敛性。但是,混沌粒子群算法在计算达到最优工期的平均时间上略比基本粒子群算法逊色。     

基于粒子群BP神经网络的质量预测模型

为了对产品质量进行预测控制、辅助新产品开发设计、寻找最优参数,将测试样本的网络输出值与真值之间的灰色关联度作为目标函数,采用粒子群算法优化了BP神经网络的权系数和阈值,构建了基于粒子群神经网络的质量预测模型。所提出的PSO-GRG算法解决了一般BP算法迭代速度慢,且易出现局部最优的问题,并以注塑件质量预测为例,进行算法实现,仿真结果表明本文所提出的PSO-GRG算法比BP算法迭代次数减少了87.5%,并避免了局部最优,且预测误差亦明显减少。得出结论:所构建的质量预测模型具有较高的预测精度和研究价值。     

求解复杂优化问题的双层多种群粒子群优化算法

为解决粒子群优化算法存在的易早熟和精度低问题,提出了一种双层多种群粒子群优化算法.此算法采用上下两层,即下层N个基础种群和上层一个精英种群.各个基础种群相互独立进化,并从精英种群中得到优良信息指导自己的进化.上层精英种群首先通过接受各基础种群的当前最优粒子来更新自己的粒子集合,然后执行自适应变异操作,最后随机地向每一个基础种群输送出本次进化后的一个最优粒子来改进其下一轮搜索.该算法的并行双进化机制增加了群体的随机性和多样性,提高了全局搜索能力和收敛精度.实例仿真表明该算法具有较好的性能,尤其对于复杂多峰函数优化,成功率显著提高.     

Optimal two-level piecewise-constant price discrimination for a storable perishable product

Price differentiation over time is an additional policy that firms might consider when determining prices for perishable products. The common policy of a fixed price regardless of freshness might result in leaving some expired inventory unsold. Price differentiation can impact the demand for perishable products, which declines as the expiration date approaches. We develop an optimisation model with the goal of evaluating the monetary effectiveness of the strategy of simultaneously combining price discrimination across heterogeneous consumers with price differentiation over time for perishable inventory under separable multiplicative demand factors of price and time. Necessary optimality equations are derived, and their solutions are proved to constitute a unique global optimal solution. It is proved that an optimal pricing policy is to implement price discrimination with respect to consumers’ sensitivity to freshness, while dynamically changing the price over time, starting with a lower price at the early stages of the product’s shelf life and increasing it at a later stage. The monetary benefit that the retailer and consumer can derive from the suggested pricing policy is evaluated by comparing the model to other models in which price discrimination or dynamic pricing are not implemented. A numerical example that illustrates the significance is introduced. From the analysis of a numerical illustration of the model, it is concluded that a dynamic price discrimination policy can be approximated by an identical-to-all dynamic pricing policy in order to maximise the retailer’s profit and thus, mitigate the retailer’s risk from failing in the process of implementing price discrimination.     

易逝品最优定价、订货量和广告投入的联合确定

以需求受价格和广告投入影响的报童问题为研究对象,需求函数采用混合形式,研究了一类随机需求下易逝品最优定价、订货量和广告投入联合确定的报童问题.建立了需求受价格和广告投入影响的报童问题的数学模型,求解得出最优定价、广告投入和订货量.并与需求仅受价格影响的报童问题做了比较,最后给出算例说明了所得的结论.     

Optimal piecewise-constant price under heterogeneous sensitivity to product freshness

When perishable products are priced uniformly, regardless of the amount of time remaining until expiration, consumers may gravitate towards fresher products, leaving some inventory unsold. This research considers dynamic pricing policies as well as replenishment policies in the context of perishable products with a fixed shelf life. Consumers are assumed to be heterogeneous in their sensitivity to freshness, i.e. their willingness to pay more for fresher products. We develop a model for identifying an optimal (profit-maximising) dynamic pricing policy and for evaluating the extent to which both the retailer and the consumer benefit from the implementation of a dynamic pricing policy as opposed to a static policy. The model assumes that the retailer is able to utilise knowledge regarding the heterogeneous characteristics of incoming customers (e.g. the retailer can gather specific information about customers’ historical purchases). Unexpectedly, it is proven that in an optimal pricing policy, the retailer should assign a lower price to fresher products and then raise the price as the products approach expiration. A numerical illustration shows that profits are strongly influenced by the volatility of consumer sensitivity to freshness; specifically, this variable has the potential to reduce optimal profits by up to 8%.     

考虑补贴水平影响的易逝品动态定价模型

建立了补贴背景下企业对易逝品动态定价的模型,并对其进行分析和求解,得出最优价格和最优期望收益关于补贴水平的一阶和二阶关系,并对企业的最优定价策略和政府对易逝品的补贴水平给出了指导性意见,得到了政府坚持公平原则的动态补贴策略。     

Agent-based simulation of pricing strategy for agri-products considering customer preference

Agri-products typically have short life cycles and are perishable. The price of perishable goods is influenced by many factors. Customers in a market have different preferences with regard to price, the quality of the product and the brand. Though studies on pricing strategies that consider customer preference are rare, this paper addresses the problem of optimal pricing strategy for retailers considering customer preferences. Traditional mathematical methods cannot adequately describe the complexities of customer preference. Due to these complexities, this paper proposes an agent-based simulation model composed of six retailers and hundreds of customers, each with personal preferences. The different retailers set prices according to freshness, inventory, cost and other factors. Due to the perishable nature of the products considered, this paper proposes a new categorising price strategy that sets prices according to different degrees of treatments. By comparing the final profit of all retailers at the end of a simulation, the categorising price strategy is demonstrated to be the optimal strategy if customers with different preferences are randomly distributed. Furthermore, based on the model, the paper studies how optimal strategies are influenced by the proportion of customers with different preferences.     

时变变质率产品的订购-定价策略

研究时变变质率产品的订购与定价联合决策问题。以系统平均利润最大化为目标函数且将已变质产品的变质处理成本考虑在内,构建了相应的订购与定价联合决策模型,其中需求同时依赖于价格和库存水平、系统允许缺货且缺货量部分延迟订购。接着,对于已知的定价策略,证明并给出了最优补货策略唯一存在的充分条件;另一方面,对于已知补货策略的情况也证明了最优定价策略的存在性。此外,证明了在联合订购与定价决策下得到的最优销售价格总是大于单独进行定价决策时得到的最优价格。进而,在模型证明和分析的基础上给出一个寻求最优解的算法。最后,通过数值模拟的方法对模型中主要参数的灵敏度进行了分析,并给出相应的管理建议。结果表明：如果零售商忽略库存水平对需求的影响或变质处理成本,制定的销售价格将偏低;对于价格弹性较高的产品,零售商应采取低价销售策略;当顾客的等待耐心较差或产品的替代性较强时,最优策略应该是提高产品的销售定价并缩短缺货期。     

Optimal markdown policy of perishable food under the consumer price fairness perception

Markdown policies are widely used by retailers to sell perishable food. Consumers purchase food at different prices during different sales periods. Some consumers may compare their experience with others who purchase the same items. Price unfairness or inequity is perceived when different prices are quoted without reasonable explanations. This study develops an optimal markdown model for perishable food pricing to optimise the food retailer revenue and enable a maximum aggregated consumer utility considering individuals’ price fairness perception. The model serves as the first step in evaluating trade-offs between food retailer revenue and consumer utility. In addition, it enables consumer utility to be depicted through perceived price fairness by including the effects of food perishability and scarcity. Another innovative feature is the equalisation of the consumer average aggregated utility during different sales periods as a condition of intertemporal price fairness perception. The proposed model is compared with two benchmark models to justify the effectiveness and advantages in the example. Finally, a sensitive analysis based on the food deterioration rate, consumer food desire rate and consumer average reservation price is conducted to justify the manner in which these factors influence the optimal pricing policy, and managerial insights are suggested for food retailers.     

Dynamic pricing vs. acquiring information on consumers’ heterogeneous sensitivity to product freshness

We propose a model of an inventory system in which a perishable product is periodically replenished, and the retailer is unaware of consumer heterogeneity in consumers’ sensitivity to freshness of a perishable product with a fixed shelf life (though it exists). Using an analytical approach, we optimally solve the problem and evaluate the extent to which unawareness is likely to detract from a retailer’s profit and the extent to which it is likely to affect the price that consumers pay. In addition, we evaluate the conditions in which a dynamic pricing policy is beneficial either to the retailer or to the consumer, as compared with a static pricing policy. It is proven that the retailer should assign products a lower price at the early stages of their shelf life and then raise the price as the products approach expiration. A numerical illustration combined with sensitivity analysis demonstrates the applicability of the modelling approach. Key parameters such as volatility of consumer sensitivity to freshness and the retailer’s estimation regarding consumers’ sensitivity are investigated. A conclusion is that, in monetary terms, it is preferable from the retailer’s perspective to invest in dynamic pricing technology rather than in technologies for gathering information on consumers’ purchases.     

基于PSO的最优投资组合计算方法

本文应用一种新型的智能算法-PSO算法求解有约束的最优投资组合问题,并讨论了最优解的质量与PSO算法中一些重要参数的关系,给出了参数选取的一些建议。数值模拟表明该算法在一定意义下优于模拟退火算法。     

供应链临时价格折扣研究

建立了一个考虑货币时间价值的供应链临时价格折扣模型,利用非线性规划分别分析了折扣生效时库存为零和不为零两种情况下零售商的定价和库存策略,并利用数值算例作了说明.研究表明,考虑货币时间价值时,传统的EOQ模型失效,而本方法能够确定不同假定下零售商的最优运营策略.