[雙語翻譯]注塑成型外文翻譯--熱塑性塑料注射成型的監(jiān)控綜述

1、3800 英文單詞， 英文單詞，2.1 萬英文字符，中文 萬英文字符，中文 6200 字文獻(xiàn)出處： 文獻(xiàn)出處：Ogorodnyk O, Martinsen K. Monitoring and control for thermoplastics injection molding a review[J]. Procedia CIRP, 2018, 67: 380-385.Monitoring and control for thermo

2、plastics injection molding A reviewOlga Ogorodnyk, Kristian MartinsenAbstractThermoplastics injection molding has found increasing use in several industry sectors. To achieve high effectivity of the process and desirable

3、 quality of the manufactured product, correct and precise parameters’ setting is critically important. As injection molding is a sophisticated process it is often hard to take care of all the changes occurring during its

4、 application. However, implementation of artificial intelligence (AI) methods in control and monitoring systems of injection molding machines can increase controllability and additivity of the process. This paper gives a

5、n overview of different studies related to research on the topic of monitoring and control systems for injection molding and explains why application of AI methods would be beneficial.Keywords: Injection molding; Artific

6、ial intelligence methods; Intelligent process monitoring and control; Intelligent heat management.1. IntroductionToday more than one third of polymeric products is produced with use of injection molding [1]. It is a comp

7、licated process, as the molten polymers undergo complex thermo- mechanical changes. As injection molding is mostly used for mass production, repeatability and quality of the final product is very important. “Improper set

8、tings of process variables will produce various defects in the final product” [2, 3] and result in increased amounts of waste and scrap. As need for control of the injection molding process is high, the first step in thi

9、s case is to precisely design, measure and monitor the process to make the key process variables observable and controllable [4]. This will allow to increase controllability and repeatability of the overall process, lead

10、ing to possibility of lowering probability of unnecessary in-process variations.The process of injection molding includes four main stages: plasticization, injection, cooling and ejection. Among these four, the cooling s

11、tage takes from 50% to 80% of the cycle time [5]. It has always been of a high interest to shorten overall cycle time, as “the cost-efficiency of the process is dependent on the time spent in the molding cycle” [6]. One

12、of the ways to shorten the cycle time and, in particular, the cooling stage, without compromising quality of manufactured parts is use of rapid heating and cooling systems, which can include application of variotherm tec

13、hnology and conformal cooling/heating channels.Process monitoring and control, as well as use of variotherm technology or conformal cooling/warming channels would benefit from application of artificial intelligence metho

14、ds in order to function in the most optimal way. The following sections will explain importance of monitoring and control systems, give examples of research on these systems and explain why AI methods are of a high impor

15、tance for the injection molding.2. Injection molding process variables and artificial intelligence methodsAccording to Karbasi and Reiser [4], the injection molding process includes three nested process loops shown in Fi

16、g. 1. The first loop called machine control includes control of machine Fig. 2. Classification of optimization methods [12]AI methods give better results when it comes to process modelling and forecasting, as they have h

17、igher precision and lower error values compared to conventional modelling methods. In addition, they are not as resource consuming as direct discrete optimization methods [12]. In order to build the model different artif

18、icial intelligence methods can be used to process big amounts of data received during the process run.Artificial neural networks (ANN) is a method that was used for modelling and forecasting in many areas of science and

19、engineering [13]. ANN is a method used for information processing, which includes use of nonlinear and interconnected processing elements called neurons. These elements are organized in separate levels connected with lay

20、ers’ weights. ANNs often consist of three layers: the input layer, the hidden layer and the output layer [14]. At first, the data is “fed” to the network’s first layer, in the second layer, it is processed and the model

21、is built, in the third layer the forecast based on the model is handed out as a result of the algorithm’s work.ANFIS or adaptive neural-based inference system is one more method used to create the models and forecasts fo

22、r certain processes. This method is a composition of artificial neural networks and fuzzy logic approaches. It identifies a set of parameters that the model will be based on using a hybrid learning rule. “It can be used

23、as a basis for constructing a set of fuzzy If-Then rules with appropriate membership functions in order to generate the previously stipulated input–output pairs” [14, 15].Genetic programming can be applied to achieve the

24、 same goals. It is a methodology which gives possibility to generate algorithms and expressions to find solution of existing problem. These expressions are represented by a tree structure consisting of leaves/terminals a

25、nd functions/nodes. When a population of the genetic programming tree is defined procedures similar to the ones used in genetic algorithm are applied. These procedures include defining the fitness function, genetic opera

26、tors (crossover, mutation and reproduction) and the termination criterion.These are only three examples of AI methods possible to be applied for injection molding, however, they need to be chosen carefully to fit a purpo

27、se of research, as well as parameters or factors used in the model.2.2. Parameters/factors influencing the quality of injection molded partQuality of injection molded part depends on a lot of factors and they are related