基于粒子群优化算法的板翅式换热器优化设计

以板翅式换热器的质量作为目标函数,以换热器芯体外形尺寸和翅片参数作为优化变量,采用粒子群优化算法对其结构尺寸进行优化设计,获得了换热器质量减轻、体积减小的效果。     

粒子群优化算法的研究

粒子群优化算法是一种基于群体智能的优化算法。介绍粒子群优化算法及其基本原理,给出各种改进技术及研究现状,并展望未来的发展趋势。     

基于改进粒子群算法的乙苯脱氢过程多目标优化

设计了一种基于支配关系构造非支配解集的多目标粒子群算法(MOPSO),将当前找到的非支配解保存到一个外部集——最优解集,利用支配更新其最优解集,多次迭代后得到Pareto最优解集。把乙苯脱氢反应过程的收率和选择性作为优化目标,动力学模型和实际生产状况作为约束条件构造乙苯脱氢过程的多目标优化问题,利用改进的多目标粒子群算法进行优化求解。基于求得的Pareto最优解集研究了各个操作条件对乙苯脱氢生产过程收率和选择性的影响,为后续乙苯催化脱氢系统实施先进控制奠定了基础。     

复合粒子群优化算法在模型参数估计中的应用

化工非线性模型的参数估计是较为困难的寻优问题，经典方法常会陷入局部极值。粒子群算法操作简便、容易实现且全局搜索功能较强，适用于非线性参数估计。但其参数值的确定与问题相关，若设定不当，会严重影响全局搜索的性能。今提出引入遗传算法，在粒子群算法的搜索过程中，逐代优选参数，包括惯性权值，加速常数，以此构建为复合粒子群优化算法。分析与测试表明，其全局搜索性能有显著改善。进一步的工作又将两种粒子群算法成功地应用于重油热解模型的参数估计。采用复合粒子群优化算法估计参数构建的重油热解模型，其预报相对误差比常规粒子群优化算法降低了8.97％，比简单遗传算法降低了23.21％，效果明显。     

粒子群优化算法在管壳式换热器设计中的应用

以成本和体积最低为设计目标,将粒子群算法用于管壳式换热器的设计过程中,为管壳式换热器的设计提供了一种新的途径.算例结果表明,该优化设计模型是可行的,并具有一定的优越性,可以应用于工程设计中.     

基于混合粒子群算法的复杂相平衡计算方法

对于含有两个部分互溶液相的相平衡问题,采用经典方法收敛困难或易陷于平凡解。为此根据最小Gibbs自由能原理,提出采用混合粒子群算法搜索全局最优解,计算得到系统的最小Gibbs自由能状态,实现复杂相平衡计算。通过改建目标函数,减少计算量,并引入组分相分率,将物料平衡约束转换为规范型立方空间的优化问题,适于粒子群算法搜索。在常规粒子群算法中引入Nelder-Mead单纯形操作,可显著提高搜优的速率和精度。将其应用于甲苯-水-苯胺液液平衡和苯-乙腈-水汽液液平衡计算,取得了良好的效果。     

基于EBF神经网络和粒子群算法的注射成型优化设计

基于拉丁超立方设计建立了椭球基（EBF）神经网络模型描述注塑工艺参数与翘曲值间的函数关系,将EBF神经网络模型与Kriging模型对比,说明EBF神经网络模型可以准确地描述注塑工艺参数与翘曲值之间的函数关系,并结合多目标粒子群算法对工艺参数进行优化,并与邻域培植遗传算法优化结果对比,说明多目标粒子群算法的优点。结果表明,基于EBF神经网络模型和粒子群优化算法可以使塑料出水管翘曲值减小11.64 %,同时使保压时间和冷却时间总和减小了2.13 s,从而在出水管批量生产过程中减少了生产时间。     

粒子群优化算法的发展及应用

讨论了粒子群优化算法的发展和应用。介绍了粒子群优化算法的基本原理和算法流程,并且与其他演化算法进行了比较,给出了一些经常用到的测试函数。针对粒子群优化算法在搜索后期存在的不足,介绍了改进的粒子群优化算法,重点介绍了在实际应用领域中用到的改进粒子群优化算法。     

基于粒子群算法的动态多目标优化

针对碳二氢生产中的反应器动态优化问题,目前虽然有多种算法对生产过程进行优化,但大部分只是对单一目标进行求解,提出一种更为灵活的反应器动态求解方法。在该方法中,首先构建碳二氢目标函数,然后采用多目标粒子群算法和分段线性函数参数法结合的方式对目标函数的进行求解,以提高整体搜索能力,得到碳二氢反应器动态优化的最优解。最后,以实际乙烯碳二加氢化工反应过程为例进行实验验证,结果证明,通过该方法进行求解的目标函数无论是在收敛性,还是在优化的平均值等方面,都比SADE-eCD和NSGA-II算法具有优势,说明该算法在反应器动态优化中是切实可行的。     

基于粒子群算法的喷水减温器建模研究

根据质量守恒和能量守恒定律建立喷水减温器的机理模型。为了提高模型精度,利用粒子群优化算法对喷水减温器的机理模型参数进行优化,并利用Matlab进行仿真,仿真结果很好地证明了模型的精确性。     

Optimal drug infusion profiles using a Particle Swarm Optimization algorithm

The dynamic optimization of the administration of therapeutic drugs in simulated patients is proposed. The approach is based on a non-linear discontinuous cardiorespiratory model, which has been conceived to simulate the effect of inotropic and vasoactive drugs as well as anesthetic agents. A stochastic technique (Particle Swarm Optimization), within the context of the control vector parameterization approach, is adopted to identify the infusion profiles of various drugs in order to track, as close as possible, the set-points on several variables of medical interest. Two different medical procedures are investigated in order to test the efficiency and robustness of the algorithm: a congestive heart failure and the unclamping of an aortic vessel. Due to the conflicting nature of the different objectives, compromise solutions are obtained in all cases.     

A Mixed Coding Scheme of a Particle Swarm Optimization and a Hybrid Genetic Algorithm with Sequential Quadratic Programming for Mixed Integer Nonlinear Programming in Common Chemical Engineering Practice

In this paper, mixed integer nonlinear programming (MINLP) is optimized by PSO_GA–SQP, the mixed coding of a particle swarm optimization (PSO), and a hybrid genetic algorithm and sequential quadratic programming (GA–SQP). The population is separated into two groups: discrete and continuous variables. The discrete variables are optimized by the adapted PSO, while the continuous variables are optimized by the GA–SQP using the discrete variable information from the adapted PSO. Therefore, the population can be set to a smaller size than usual to obtain a global solution. The proposed PSO_GA–SQP algorithm is verified using various MINLP problems including the designing of retrofit heat exchanger networks. The fitness values of the tested problems are able to reach the global optimum.     

Optimization of experimental conditions based on Taguchi robust design for the preparation of nano-sized TiO2 particles by solution combustion method

This investigation was aimed at preparing nanocrystalline TiO₂ powder by solution combustion method, and searching the optimum preparing conditions by employing Taguchi robust design method. Taguchi robust design method with L₁₈ orthogonal array was implemented to optimize experimental conditions for preparing nano-sized titania particles. Titanium IV n-butoxide was hydrolyzed to obtain titany1 hydroxide [TiO(OH)₂], and titanyl nitrate [TiO(NO₃)₂] was obtained by reaction of TiO(OH)₂ with nitric acid. Finally, the aqueous solution containing titanyl nitrate [TiO(NO₃)₂] and a fuel, glycine, were mixed and combusted to obtain the nano-sized titania. The optimum conditions obtained by this method are as follows (based on 1 mol of TiO₂ per batch): concentration of HPC, 0.053 mg cm⁻³; mole ratio of Ti:H₂O:IPA, 1:4:10; hydrolysis time, one hr; the amounts of HNO₃ and glycine are 10 ml and 0.5 g, respectively; nitrated temperature, 298 K and nitrated time, 2 h. TiO₂ nanocrystalline (∼15 nm) with high BET surface area (350 m² g⁻¹) and narrow band gap energy (2.7 eV) were thus obtained.