共查询到19条相似文献,搜索用时 250 毫秒
1.
采用真空蒸馏法,以丙酮为溶剂,通过改变蒸馏温度、溶液初始质量浓度、系统真空度等参数,探讨了各因素对NTO重结晶的粒度分布及晶型的影响.对实验结果进行理论分析,确定了最佳重结晶细化NTO的工艺参数.结果表明,晶型主要受蒸馏温度的影响,其次是系统真空度.在蒸馏温度60 ℃左右、系统真空度0.05 MPa~0.06 MPa、溶液初始质量浓度不低于0.2 g/12 mL时,通过真空蒸馏法可以得到比较规则的立方晶型,颗粒粒度在1 靘左右,且分布范围窄. 相似文献
2.
由于提金工艺参数的优化对提金率的提高和原料的高效利用有至关重要的影响,因此基于数据挖掘技术和关联规则理论,采用Apriori算法挖掘了提金工艺历史数据库中对提金率有影响的可控参数关联规则,得到了可控过程参数间的相对关系曲线,确定了工艺参数优化区间,有效提高了提金率。可为类似的流程工业生产优化问题的研究提供参考。 相似文献
3.
提出用免疫量子粒子群算法优化控制决策表,使控制决策表的参数整定简单易行.其核心思想是将控制决策表作为算法中的粒子,以迭代搜索的方式寻找全局最优粒子.该算法的全局寻优能力强,计算机实现简单,可调参数少.模糊控制器和控制决策表的优化设计在SCON-2000模糊控制平台进行了工程化实现,并对水箱液位进行模糊控制.从对比结果中可以看出,优化了控制决策表以后,系统响应更快,精度更高,抗扰动能力更强.这表明了该算法在模糊控制器参数优化中的可行性. 相似文献
4.
5.
6.
通过利用参数整定PID模糊控制器对LPG换热系统中被加热介质丙烷的出口温度进行控制。根据其动态参数和相关经验制定模糊规则,利用模糊规则对PID三个参数Kp、Ki、Kd进行调整,以改善其控制性能,应用MATLAB软件对实际系统进行仿真。结果表明,该控制方法具有超调量小、过渡时间短、适应性强、抗干扰性强等优点。最后,对照实际运行结果验证了模糊PID控制的优越性。 相似文献
7.
8.
基于两参数状态方程提出的LCVM混合规则是将以无穷压力为参考态的HV混合规则与以零压力为参考态的MHV1混合规则线性结合而得的,尽管没有理论基础,但对于计算非极性、极性体系给出了较好的结果.因而研究借鉴获得LCVM混合规则的思路将LCVM混合规则扩展到多参数状态方程中,即将HVOS混合规则与MHV1混合规则进行线性结合,并通过引入两个对比参数λ0,λ∞给出了适合多常数状态方程的新LCVM型混合规则.在新的LCVM型混合规则中,参数δ决定了MHV1和HVOS混合规则的相对贡献,该参数可由模型在高压和低压下拟合二元混合物的泡露点得到,通过拟合得到的值约为0.21;混合规则中的活度系数模型可以利用由低压区关联出的GE模型.采用该新混合规则模型,在较宽的温度与压力范围内,结合Harmens-Knapp(HK)方程对包括非极性体系、极性体系等在内的20种二元混合物进行了相平衡计算.计算的结果与实验数据吻合得很好.该模型与采用VDW混合规则模型的相平衡计算结果比较表明,该模型的关联精度有了很大的提高,可以在较大的温度与压力范围内关联多种体系的汽液相平衡数据.但是在混合规则中,参数b,c是依赖于经验获得,因此还有待于进一步的研究. 相似文献
9.
10.
11.
12.
13.
将粗集理论应用到旋转机械故障的智能诊断中,在信息不完备情况下,不需要事先的假设和专家经验,在 粗糙度不变的情况下进行推理和在决策分明性不变意义下进行决策,找出其规律性,并提供规则的可信度;同时能 对旋转机械故障的知识库的进行约简,消除多余的条件属性和冗余的信息,导出最小简化决策表。 相似文献
14.
15.
针对非线性动态系统的控制问题,提出了一种基于自适应模糊神经网络(adaptive fuzzy neural network, AFNN)的模型预测控制(model predictive control, MPC)方法。首先,在离线建模阶段,AFNN采用规则自分裂技术产生初始模糊规则,采用改进的自适应LM学习算法优化网络参数;然后,在实时控制过程,AFNN根据系统输出和预测输出之间的误差调整网络参数,从而为MPC提供一个精确的预测模型;进一步,AFNN-MPC利用带有自适应学习率的梯度下降寻优算法求解优化问题,在线获取非线性控制量,并将其作用到动态系统实施控制。此外,给出了AFNN-MPC的收敛性和稳定性证明,以保证其在实际工程中的成功应用。最后,利用数值仿真和双CSTR过程进行实验验证。结果表明,AFNN-MPC能够取得优越的控制性能。 相似文献
16.
《Journal of Adhesion Science and Technology》2013,27(6):635-654
A major problem in industrial applications of structural adhesives is the quality assurance of manufactured joints. At present, for lack of a suitable nondestructive technique, production standards for adhesively-bonded aluminum joints are established on the basis of destructive tests and statistical inference. An experimental study was carried out to assess if lock-in thermography (LT) could be used as a tool for nondestructive evaluation of adhesively-bonded aluminum joints. Several samples were fabricated by varying the governing parameters such as nature of aluminum alloy, substrate thickness, surface treatment, adhesive type and bondline thickness. The effects of surface treatments on the loading capability of lap joints were evaluated through both destructive tensile tests and nondestructive evaluation with infrared LT. Tensile tests showed that the joint performance was not affected by the nature of the aluminum alloy but by the substrate thickness, the adhesive type and the bondline thickness. LT was capable of detecting imperfections such as scratches on substrates and foreign inclusions in the adhesive layer. 相似文献
17.
18.
粗糙集与模糊推理相集成的过程建模方法及其应用 总被引:1,自引:0,他引:1
针对复杂化工过程机理建模困难的问题,采用适应性较广的模糊方法经验建模,鉴于模糊法对于高维、强相关的样本数据很难导出规则,本文提出先用粗糙集方法消除冗余性,约简系统,获取最小规则集,在此基础上构建结构合理、参数可适当初始化的模糊-神经网络,并采用LM算法训练,收敛速率快,模型预测性能良好.将此法用于PTA装置溶剂脱水塔精馏过程的经验建模,效果令人满意,性能优于现代统计方法和前馈神经网络. 相似文献
19.
《Journal of Adhesion Science and Technology》2013,27(7):589-632
Infrared thermography (IRT) is a non-contact, non-intrusive technique which converts the invisible thermal energy, radiated from the surface of an object in the infrared band of the electromagnetic spectrum, into a video signal, each energy level being generally represented by a color or a gray level. IRT has been considered an exciting scientific breakthrough since its introduction in the early 1960s. Indeed, the new generation of fully-computerized infrared imaging systems can provide both qualitative and quantitative measurements which are useful in many industrial and research fields. Nevertheless, infrared thermography is still not completely exploited. The reason for this lies, in part, in the lack of knowledge, since at first sight IRT seems to be too expensive and difficult to use and, in part, in the industrial inertia to change the routine procedures. The aim of this review article is to provide the reader with a background to infrared theory and with an overview of the most relevant applications of IRT to the adhesion field. The use of IRT as a non-destructive evaluation technique with the two different approaches of pulse thermography (PT) and lock-in thermography (LT) is discussed. Many applications are described which involve several different materials (metals, plastics, plaster, composites, hybrid composites and sandwiches) and different types of bonds (coatings, sandwiches and joints). The results show that both PT and LT are able to detect material modifications caused by surface treatments, presence of inhomogeneities in bulk materials, as well as disbonding, delamination, and cracks and slag inclusions in bonded structures. The LT is also capable of evaluating materials characteristics (e.g., variations in density, porosity, hardness, etc., which induce variations in the phase angle), the dimension of the heat affected zone in welded joints, coating thickness, bondline thickness, the effects of adhesive thickness, the effects induced in bonded structures by substrate surface treatments, and the effects of crosslinking in polymers. The LT technique is particularly advantageous in the evaluation of frescoes, mosaics and antique artworks. The reported applications provide also information which is useful for decision making about the use of IRT alone, or combined with other techniques. 相似文献