دانلود مقاله مدل کسب و کار تولید هوشمند با اینترنت اشیا و یادگیری ماشینی
چکیده
1. مقدمه
2. مرور مطالعات پیشین
3. مدل کسب و کار تولید هوشمند تحت محیط اینترنت اشیا و یادگیری ماشینی
4. تحلیل امکان سنجی الگوریتم شبکه عصبی هوشمند مصنوعی در مدل کسب و کار تولید هوشمند
5. نتیجه گیری
منابع
Abstract
1. Introduction
2. Literature review
3. The business model of intelligent manufacturing under the environment of the Internet of Things and machine learning
4. The feasibility analysis of artificial intelligent neural network algorithm in the business model of intelligent manufacturing
5. Conclusions
Disclosure statement
References
چکیده
برای ایجاد یک مدل کسب‌وکار تولید هوشمند، از شبکه متخاصم مولد (SeqGAN) برای بهینه‌سازی الگوریتم شبکه عصبی Back Propagation (BP) که توسط الگوریتم ژنتیک چند هدفه بهبود یافته استفاده شد تا دنباله شبکه متخاصم مولد- الگوریتم الگوریتم ژنتیکی را ارائه دهد. (SeqGAN-GABP). در همین حال، الگوریتم Elman توسط مدل SeqGAN برای پیشنهاد الگوریتم SeqGAN-Elman بهینه شد. الگوریتم ها ساخته و آموزش داده شدند و در پلتفرم های اینترنت اشیا اعمال شدند. نتایج نشان داد که الگوریتم SeqGAN-GABP از نظر حداقل خطا، دقت برازش، زمان آموزش و استفاده از حافظه داخلی از الگوریتم SeqGAN-Elman بهتر عمل می کند.
توجه! این متن ترجمه ماشینی بوده و توسط مترجمین ای ترجمه، ترجمه نشده است.
Abstract
To establish a business model of intelligent manufacturing, the sequence Generative Adversarial Network (SeqGAN) was used to optimise the Back Propagation (BP) neural network algorithm improved by multi-objective Genetic Algorithm to propose the sequence Generative Adversarial Network-Genetic Algorithm Back Propagation Algorithm (SeqGAN-GABP). Meanwhile, the Elman algorithm was optimised by the SeqGAN model to propose the SeqGAN-Elman algorithm. The algorithms were constructed and trained and were applied to the Internet of Things platforms. The results showed that the SeqGAN-GABP algorithm outperforms the SeqGAN-Elman algorithm in terms of minimal error, fitting accuracy, training time and internal memory usage.
Introduction
At present, the domestic and foreign manufacture industries are moving towards an intelligent and digital era, and the influence of intelligent manufacturing on various aspects of manufacture industry is also growing (Peng and Gao 2017; Mousavi et al. 2017). Doubtlessly, it is the developing direction of the automatic manufacture for intelligent manufacturing. The intelligent manufacturing system judges and plans its own behaviour by collecting and analysing its own information and environmental information and enriches the knowledge base in the practice process (Li et al. 2018). Business model refers to various transaction relationships and connection methods between enterprises and enterprises, between departments and departments, between enterprises and customers, and between enterprises and channels (Kaulio, Thorén, and Rohrbeck 2017). A business model is a conceptual tool that contains a set of elements and their relationships to illustrate the business logic of a particular entity (Ding et al. 2017). It describes the value that a company can provide to its customers, as well as the internal structure, partner network, and relationship capital of the company to achieve (create, market, and deliver) this value and generate sustainable profits (Yun, Won, and Park 2017).
Conclusions
Through the business model of intelligent manufacturing based on IoT and machine learning, the artificial neural network algorithm and the IoT platform have reliability in the business model of intelligent manufacturing, which could improve the development of the business model of intelligent manufacturing, facilitate the user interaction and business development, and have broad application prospects. However, in the actual application process, relevant analysts should cooperate and coordinate based on the actual problems to clarify the analysis objectives and feasibility of machine learning. However, there were also deficiencies in the research process. For example, the machine learning model established in this study needs to manually evaluate the underlying indicators. In the subsequent research, the underlying indicators can be judged by computer vision, thereby gradually eliminating the stage of manual data input and evaluation. The scale of the network scale fails to adapt to the background of big data. In the future, the feature vector of the evaluation system model can be raised to high-dimensional space, while the high-dimensional low-level features are extracted through the deep learning network and compressed to low-dimensional advanced features. Therefore, network performance needs to be improved through machine learning. In addition, the innovation and reconstruction of business models involve all aspects of the industry. The distribution of interest by different companies and the better integration of industry resources are also a major obstacle to the implementation of an industrial blockchain.