Scholarlink

<p>The Smart home system is an advanced technological development that highly assists with energy management, security, and ease of access. However, the smart devices gave inappropriate notifications, which are susceptible to spoofing and masking attacks, resulting in the introduction of Prediction Maintenance strategies. In conventional prediction maintenance techniques, the lack of appropriate thermal response characteristics led to a drift in the sensitivity of the sensor, causing false alarms, which brought dissatisfaction and a decrease in the efficiency of the system. Hence, in the proposed work, an Intelligent prediction maintenance strategy with optimized energy management is introduced for Smart home applications. The proposed work uses a Novel Flexi Neuro Temporal Fuzzy Controller (FNTFC), which uses Bilateral-Long Short Term Memory (Bi-LSTM) for regulating the temporal dynamics of the sensor data. To capture intricate interactions, behaviors, and dependencies over sensors of the neighborhood, a Non-linear Temporal Neighbourhood Fuzzy Logic Controller (NFLC) along with Temporal Convolutional Network (TCN) and Time Decay Mechanism (TCM) is employed. Besides, to attain efficient dynamic load balancing, a Novel Wavelet Convo Decomposer is used, which brings together a CoifVarNet and a Convolutional Variational Autoencoder for monitoring the real-time efficiency variation and to compute the efficiency of sensor data features. Then, for optimizing the energy consumption, a Seasonal Tabu Decomposed Optimization is used, which thereby brings dynamic load balancing. The proposed work is implemented in the MATLAB platform, for which a smart home dataset with 500,000 readings corresponding to various equipment is considered, from which it is concluded that the defined work gives high Accuracy (96.9%), Precision (98%) and Energy efficiency (98.3%), thereby enhances predictive maintenance and energy optimization. The proposed work, when combined with Renewable energy sources, assists in practical applications like Electric Vehicles and Smart grids.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Predictive maintenance and optimized energy management in smart homes using machine learning based intelligent controller

  • 期刊论文

  • 发表日期：

    2025年12月27日

查看全文
  • Abhijeet R Raipurkar,
  • Manoj B Chandak,
  • Sunita G Rawat

摘要

The Smart home system is an advanced technological development that highly assists with energy management, security, and ease of access. However, the smart devices gave inappropriate notifications, which are susceptible to spoofing and masking attacks, resulting in the introduction of Prediction Maintenance strategies. In conventional prediction maintenance techniques, the lack of appropriate thermal response characteristics led to a drift in the sensitivity of the sensor, causing false alarms, which brought dissatisfaction and a decrease in the efficiency of the system. Hence, in the proposed work, an Intelligent prediction maintenance strategy with optimized energy management is introduced for Smart home applications. The proposed work uses a Novel Flexi Neuro Temporal Fuzzy Controller (FNTFC), which uses Bilateral-Long Short Term Memory (Bi-LSTM) for regulating the temporal dynamics of the sensor data. To capture intricate interactions, behaviors, and dependencies over sensors of the neighborhood, a Non-linear Temporal Neighbourhood Fuzzy Logic Controller (NFLC) along with Temporal Convolutional Network (TCN) and Time Decay Mechanism (TCM) is employed. Besides, to attain efficient dynamic load balancing, a Novel Wavelet Convo Decomposer is used, which brings together a CoifVarNet and a Convolutional Variational Autoencoder for monitoring the real-time efficiency variation and to compute the efficiency of sensor data features. Then, for optimizing the energy consumption, a Seasonal Tabu Decomposed Optimization is used, which thereby brings dynamic load balancing. The proposed work is implemented in the MATLAB platform, for which a smart home dataset with 500,000 readings corresponding to various equipment is considered, from which it is concluded that the defined work gives high Accuracy (96.9%), Precision (98%) and Energy efficiency (98.3%), thereby enhances predictive maintenance and energy optimization. The proposed work, when combined with Renewable energy sources, assists in practical applications like Electric Vehicles and Smart grids.