دانلود مقاله مدل پیش‌بینی شبکه عصبی هوش مصنوعی برای پایش توربین های بادی با داده های SCADA
چکیده
مقدمه
روش ها
نتایج
بحث
نتیجه گیری
منابع
Abstract
Introduction
Methods
Results
Discussion
Conclusion
References
چکیده
صنعت 4.0 یک تغییر پارادایم در نحوه نظارت و تشخیص تجهیزات صنعتی با کمک فناوری های نوظهور مانند هوش مصنوعی (AI) ایجاد کرده است. ابزارهای عیب‌یابی مبتنی بر هوش مصنوعی نقش مهمی در فرآیندهای تشخیص و نظارت با کارایی بالا، به‌ویژه برای سیستم‌های متشکل از چندین مؤلفه از جمله توربین‌های بادی (WT) دارند. استفاده از چنین رویکردهایی نه تنها زمان عیب یابی و تشخیص را کاهش می دهد، بلکه با پیش بینی رفتار اجزای مختلف و محاسبه احتمال خرابی در آینده نزدیک، پیشگیری از خطا را ممکن می سازد. این نه تنها هزینه های تعمیر را با ارائه نظارت ثابت اجزا و شناسایی علل عیوب کاهش می دهد، بلکه با کاهش زمان های خرابی به دلیل سیستم هشدار اولیه مبتنی بر هوش مصنوعی، کارایی دستگاه را نیز افزایش می دهد. این مقاله هشت مدل مختلف شبکه عصبی مصنوعی (ANN) را برای تشخیص و نظارت بر WTs ارزیابی، مقایسه و مقایسه کرد که خرابی سیستم ماشین‌آلات را بر اساس سیگنال‌های حسگر اجزای داخلی و دمای تولید پیش‌بینی می‌کنند. این مقاله از یک رویکرد مدل یادگیری ماشین با دو لایه پنهان با استفاده از رگرسیون خطی چندلایه برای دستیابی به هدف خود استفاده کرد. سیستم توسعه‌یافته خروجی دمای ژنراتور WT را با دقت ۹۹.۸ درصد با پیش‌بینی اندازه‌گیری ۲ ماهه پیش‌بینی کرد.
توجه! این متن ترجمه ماشینی بوده و توسط مترجمین ای ترجمه، ترجمه نشده است.
Abstract
The industry 4.0 has created a paradigm shift in how industrial equipment could be monitored and diagnosed with the help of emerging technologies such as artificial intelligence (AI). AI-driven troubleshooting tools play an important role in high-efficacy diagnosis and monitoring processes, especially for systems consisting of several components including wind turbines (WTs). The utilization of such approaches not only reduces the troubleshooting and diagnosis time but also enables fault prevention by predicting the behavior of different components and calculating the probability of near future failure. This not only decreases the costs of repair by providing constant component’s monitoring and identifying faults’ causes but also increases the efficacy of the apparatus by lowering the downtimes due to the AI-driven early warning system. This article evaluated, compared, and contrasted eight different artificial neural network (ANN) models for diagnosis and monitoring of WTs that predict the machinery’s system failure based on internal components’ sensor signals and generation temperature. This article employed a machine learning model approach with two hidden layers using multilayer linear regression to achieve its objective. The developed system predicted the output of the WT’s generator temperature with an accuracy of 99.8% with 2 months in advance measurement prediction.
Introduction
Industry 4.0 introduced a new paradigm to the machineries’ monitoring and diagnosis procedure. Enabled by advances in artificial intelligence (AI) and notions such as Internet of Things (IoT) in recent years, Industry 4.0 has proved to be an effective and reliable trend toward digitization and automation (Haag and Anderl 2018). Industry 4.0 is the fourth paradigm shift and major breakthrough in industrial revolution made possible by the advancements in electronics and information technology; a continuation of the evolution of automation commenced from the invention of steam engines and mass production as a result of assembly lines and standardization (Xu, Xu, and Li 2018).
The wind power industry and the whole renewable energy sector could benefit significantly from the employment of industry 4.0. Most of the machineries including wind turbines (WTs) produce a huge amount of data related to power consumption, current, voltage, vibration, and environmental factors that are not necessarily utilized. The information processed from these gathered data could be used to improve the troubleshooting, monitoring, and maintenance procedures. Sensors attached to different parts of WTs will provide important data of the health state of the apparatus, which require interpretation and processing.
Conclusion
Eight different ML models were developed with different sensors’ data based on SCADA data collected from nine WTs over 10 years received from the Westmill Windfarm located in Swindon, United Kingdom, to predict the generator failure in WTs in advance using pattern recognition based on historical data. The results of each model’s accuracy in terms of minimum, maximum, and standard deviation offsets between the predicted and actual generator temperature values were compared and contrasted, and the effect of the input sensor data was explored. Overall, this research showed the possibility of utilizing ML-driven regression algorithms to predict WTs’ generator failure caused by heat, lowering the maintenance costs related to downtime and staff,and, at the same time, improving the operational availability of the apparatuses. For the future works, the authors of this article aim to explore the possibility of implementing transfer learning for fast adaptation and deployment of the trained models to new WTs, allowing quick training of new assets and lowering the readiness time required for the model.