Figure 5 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Dataset preparation procedure. 1) Data collection: Environmental data read from sensors are transmitted to the cloud IoT Hub infrastructure and stored in the database. 2) Dataset preparation: The raw dataset is analyzed and performed data cleaning. It is then split into training set and test set. Anomaly generation is applied to the test set to create point and context anomaly datasets. 3) Feature preparation: Temporal features and prediction, sliding window or overlapped sliding window dependent features are extracted from training, test and anomaly datasets to create the final training and test datasets for different detection methods.
Source publication
![Figure 3: The architecture of an LSTM-encoder-decoder [63]](publication/343846811/figure/fig2/AS:928238398341122@1598320688248/The-architecture-of-an-LSTM-encoder-decoder-63_Q320.jpg)
Indoor climate is closely related to human health, comfort and productivity. Vertical plant wall systems, embedded with sensors and actuators, have become a promising application for indoor climate control. In this study, we explore the possibility of applying machine learning based anomaly detection methods to vertical plant wall systems so as to...
Contexts in source publication
Context 1
... this study, two different categories of anomaly detection methods, i.e., prediction-based and pattern recognitionbased detection methods, are applied to the sensory data collected from an indoor environment, to examine the effectiveness of detecting both point anomalies and contextual anomalies, and verify the possibility to integrate real-time anomaly detection into the cloud-based plant wall system. In this section, we demonstrate our experiment details. Figure 5 illustrates how the training and test datasets are constructed. In the experiment, the raw data is based on the collected CO 2 data by a vertical plant wall that has been placed in an elderly home located in Norrköping, Sweden . ...
Similar publications
The effect of Artificial Intelligence (AI) on smart manufacturing is increasing rapidly. The AI-based open-source web application enables users to build and distribute live code, calculations, visualizations, and explanatory text documents, and also supports the data cleaning and conversion, numerical simulation, mathematical modelling, machine lea...
Lakehouse systems have reached in the past few years unprecedented size and heterogeneity and have been embraced by many industry players. However, they are often difficult to use as they lack the declarative language and optimization possibilities of relational engines. This paper introduces RumbleML, a high-level, declarative library integrated i...
Citations
... Proactive measures similar to predictive maintenance are essential to maximize efficiency and comfort. Identifying early anomalies in indoor climatic parameters and intervening promptly to improve occupant well-being and decrease energy consumption is crucial [11]. Following the detection of attacks, implementing control strategies such as PID (Proportional-Integral-Derivative), Fuzzy Logic, and Model Predictive Control (MPC) can facilitate the system in taking appropriate actions to mitigate the impact of the attacks. ...
With the advent of IoT technology, smart building management has been transformed, leading to significant improvements in energy efficiency and occupant comfort. Indoor room temperature control is crucial as it affects both building performance and occupant quality of life. Nevertheless, stringent cybersecurity measures are required due to the increasing susceptibility to cyber attacks with more IoT links in smart buildings. Identifying and managing unusual temperature readings is essential to keep the system running smoothly, efficiently, and safely. By integrating classical control methods such as PID with anomaly detection and LSTM modeling, this approach enables proactive anomaly identification and accurate temperature forecasts , rendering sustainable and resilient living conditions. This integration optimizes resource usage and mitigates cyber risks. This paper presents a holistic method that combines PID control, LSTM forecasting, and anomaly detection for smart building applications. The proposed integrated approach successfully addresses aberrant temperature variations and enhances building performance, as shown through experimental validation.
... This layout was developed with the intention of reducing the inaccuracy brought on by data scattering and increasing precision. In contrast, Liu, et al. [36] established a model for detecting point abnormalities and contextual anomalies in vertical plants by combining AE with long short-term memory (LSTM). Each of the two LSTM networks that comprise this model-the encoder and the decoder-plays distinct roles. ...
Devices connected to the Internet of Things are expanding quickly and producing tons of data. Simultaneously, malicious attempts to gain access to sensitive data or disrupt networks have accelerated and grown more sophisticated. As a result, cybersecurity has emerged as a critical matter in the evolution of prospective networks capable of responding to and countering such threats. Intrusions detecting systems (IDS) are crucial for IoT security. There are two types of intrusion detection systems: signature-based and anomaly-based. Signature based work to identify known attacks. However, it will be unrealistic to rely on pre-defined threat monitoring (signature-based) due to the variety of attack types in addition to the unstructured data generated by IoT devices and their characteristics. On the other hand, anomaly-based intrusion detection is able to identify known and unknown attacks. Many anomaly-based systems were designed using autoencoder techniques to reduce the high diminution of data and identify the intrusion in the resource-constrained IoT devices. This paper proposed an in-depth review of the approaches that used an autoencoder to detect attacks in an IoT environment. The main objective of this review is to find and address the current challenges of anomaly intrusion detection systems based on autoencoders as well as to present an overview of the different types of autoencoders that are used to enhance the effectiveness of intrusion detection systems. The findings of this paper show that autoencoders achieved an encouraging result in terms of reducing the dimensionality, improving the accuracy, and detecting anomalies. However, these approaches still suffer from high training time, complexity, and detection time. A robust adaptive anomaly intrusion detection system based on an autoencoder, to reduce the high dimensionality of data and detect attacks in IoT, is much needed.
... The continuous probability flood hazard maps obtained from the ML models are converted into binary flood presence-absence maps that highlight the flood hazard presence extent. To achieve this, a threshold selection method based on the closest point to (1,0) on the Receiver Operating Characteristic (ROC) curve is employed (Huynh et al. 2015;Liu et al. 2020). By analyzing the ROC curve generated from the classification model's predictions, the threshold that corresponds to the operating point nearest to (1,0) is determined. ...
Floods have become increasingly frequent and devastating in recent decades, posing unignorable risks as highly destructive natural hazards. To effectively manage and mitigate these risks, accurate flood hazard mapping is crucial. Machine learning models have emerged as valuable approaches for flood hazard assessment. In this study, six machine learning (ML) models, including Maximum Entropy, Support Vector Machine, Extreme Gradient Boosting (XGB), Random Forest (RF), multi-layer perceptron, and Naive Bayes, were utilized to evaluate urban flood hazard in Zaio, NE Morocco, and estimate the flood presence extent. Nine flood conditioning factors were used as input variables. Historical flood presence and absence data were employed for models training and testing, incorporating 663 flood presence and absence locations dating back to past flood events. Performance evaluation metrics such as Kappa statistic, accuracy, sensitivity, specificity, and area under the curve (AUC) were calculated for each model. RF (AUC = 0.92) and XGB (AUC = 0.9) models showed excellent classification capabilities, surpassing the performance of the other models, while the other models exhibited lower but recognizable performances. Additionally, the hazard presence extent maps generated by the ML models exhibited a decent alignment with a historical flood event maps created by the hydrodynamic and the cellular automata models. The results imply that ML models offer effective solutions for mapping urban flood hazards. The innovative integration of various ensemble and single ML models demonstrates their potential in urban flood hazard susceptibility and extent mapping, effectively surpassing the limitations associated with limited availability of hydrologic/hydraulic data and computational burden. These mapped results can be instrumental for local authorities in shaping mitigation strategies in the city of Zaio.
... Typically, higher prerequisites of data, algorithm, and virtualisation are required by prediction than detection, resulting in less failure prediction-related research. In the IAP field, many existing solutions have been developed for status monitoring and anomaly detection, among which IoT and artificial intelligence (AI) algorithms are demonstrated as utility and efficiency solutions [6][7][8]. Therefore, leveraging the collected environmental data to achieve failure prediction with advanced algorithms and improved interaction platforms holds great promise and value. In this study, we proposed a paralleled deep learning (DL) model for the failure prediction of IAQ, aiming to enhance the automation and reliability of PdM of indoor climate. ...
... Vertical green living walls (VGW) are recognised as a nature-centric strategy to enhance air quality and enhance comfort within the construction sector. Liu et al. [6] explored the possibility of deploying ML-based anomaly detection methods to vertical plant wall systems to enhance automation and improve intelligence to realise the PdM of the indoor climate. Besides, Helman et al. [35] developed a comprehensive high-throughput monitoring system utilising remote sensing technology in a VGW workplace, which integrates multiple data acquirement systems such as sensor systems, cameras, gas-exchange measurements, and AI algorithms into an analytical and visualisation of an online platform. ...
... 3) Output stage: Following Eq. (5), the LSTM determines the current output from the cell state using the output gate (z o ), and a new hidden layer state h t is created with the new cell state c t as shown in Eq. (6). The output y t in LSTM is typically obtained by applying a transformation to the h t like regular RNNs, illustrate in Eq. (7). ...
... The indoor climate in contemporary buildings, encompassing air temperature, relative humidity, CO 2 concentration, fine dust concentration, illuminance, and total volatile organic compounds (TVOCs), plays a pivotal role in human health, comfort, and task productivity [1][2][3][4]. The quest for techniques that can effectively conserve energy while upholding clean and pleasurable indoor environments is of paramount importance and has garnered significant attention from both society and academia [1,2,5]. ...
... The indoor climate in contemporary buildings, encompassing air temperature, relative humidity, CO 2 concentration, fine dust concentration, illuminance, and total volatile organic compounds (TVOCs), plays a pivotal role in human health, comfort, and task productivity [1][2][3][4]. The quest for techniques that can effectively conserve energy while upholding clean and pleasurable indoor environments is of paramount importance and has garnered significant attention from both society and academia [1,2,5]. The continuous evolution of computing and machine learning technologies has significantly contributed to optimizing human comfort and health, as well as reducing energy consumption [6,7]. ...
This study focuses on gathering environmental data concerning the indoor climate within a dormitory, encompassing variables such as air temperature, relative humidity, CO2 concentration, fine dust concentration, illuminance, and total volatile organic compounds. Subsequently, an anomaly detection long short-term memory model (LSTM) model, utilizing a two-stacked LSTM model, was developed and trained to enhance indoor environment control. The study demonstrated that the trained model effectively identified anomalies within eight environmental variables. Graphical representations illustrate the model’s accuracy in anomaly detection. The trained model has the capacity to monitor indoor environmental data collected and transmitted using an Internet-of-Things sensor. In the event of an anomaly domain prediction, it proactively alerts the building manager, facilitating timely indoor environment control. Furthermore, the model can be seamlessly integrated into indoor environment control systems to actively detect anomalies, thereby contributing to the automation of control processes.
... However, due to the constraints of the empirical models' structures and relatively simple algorithms, the accuracy and generalization ability of regression models is rather limited (Wu et al., 2022). Fortunately, machine learning technology with multi-dimensional collaborative analysis capabilities has been introduced into thermal environment simulation research (Liu et al., 2021;Liu et al., 2020). Including the artificial neural network (ANN), support vector machine (SVM), and random forest (RF), each model has various advantages (Abdulla-Al et al., 2022). ...
Industrial areas emit large amounts of extremely high temperatures and aggravate the extreme urban heat island (E-UHI) effect, which has become a non-negligible environmental issue worldwide. Due to the lack of in-depth understanding of the relationship between E-UHI and surface structural synergy, few studies explored the causes and mitigation measures of E-UHI. Therefore, this study accordingly defined the E-UHI effect at a finer scale based on the thermal characteristics and human thermal perceptions, and a multi-indicator' method based on machine learning was proposed to quantify internal causes and collaborative mitigation measures. The results revealed that the average E-UHI scale shows a quadratic growth in Wu'an. Artificial heat from production space (PS) is the main cause of E-UHI formation. The synergy between green space (GS), PS and shape of industrial land determines the development of E-UHI. Although both the decrease of PS and the increase of GS are ways to mitigate high land surface temperature, when PS is higher than 50%, the cooling effect of GS is not obvious. When GS is less than 2.5%, the adjustment of PS is invalid for improving the thermal environment. For mitigating E-UHI, a factory with high PS needs to increase its shape index. The synergy between the structural indicators is a novel model to solve the shortcoming of adjusting a single factor in mitigating complex E-UHI. The results provide theoretical guidance for mitigating E-UHIs.
... The authors proposed a method for detecting specific anomalies in energy consumption that can be applied to smart buildings, as reported in their publication [9]. The authors of reference [15] proposed an unsupervised dynamic threshold technique in detecting anomalies in time-series data related to temperature. Wang et al. [16] proposed an improved long short-term memory (LSTM) model for forecasting energy consumption. ...
Energy efficiency in modern homes has recently become a significant issue due to the emergence of smart home infrastructure. Numerous public structures, such as homes, hospitals, schools, and other institutions, use more energy. To come close to meeting the actual energy demand, it is crucial that we create as much energy as we can. Machine learning has various advantages for improving the effectiveness and efficiency of smart home systems and appliances, including managing and lowering energy use. Additionally, as a key component of the smart home idea, we explore the potential integration of machine learning-based on some algorithm methodologies ways to improve power energy consumption system and control. The models were used to identify patterns for smart home and variations in energy consumption. This study's conclusions were used to analyze case studies and forecast energy consumption. Detection Change (of used and generation) for all appliances, which excessive foresees energy use and stops a rise in usage. Predict Future Energy use by using meteorological data and maximizing the supply of energy to forecast future energy generation and use. Finally, using five machine learning algorithms, including the Linear Regression (LR), Gradient Boosting Regression (GBoostR), Decision Tree Regression (DTR), Stochastic Gradient Descent Regression (SGDR), and Bayesian Ridge Regression (BRR), we can measure the Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Absolute Error (RMAE), and Root Mean Squared Percentage Error (RMSPE), in order to determine how well models.
... where w k t ∼ N µ k , σ 2 k represents the gas concentration present in normal operating conditions in the surrounding of the kth sensor [37], where the values of µ k and σ 2 k are both known. The values of {µ k } k and σ 2 k k can be estimated by calculating the sample mean and sample variance from a set of measurements acquired in normal operating conditions (H 0 ). ...
... We here report the derivation of the ATP CSA reported in Eq. (37). We show the proof for the case where H 0 is true: ...
This work addresses the problem of detecting gas dispersions through concentration sensors with wireless transmission capabilities organized as a distributed Wireless Sensor Network (WSN). The concentration sensors in the WSN perform local sequential detection (SD) and transmit their individual decisions to the Fusion Center (FC) according to a transmission rule designed to meet the low-energy requirements of a wireless setup. The FC receives the transmissions sent by the sensors and makes a more reliable global decision by employing a SD algorithm. Two variants of the SD algorithm named Continuous Sampling Algorithm (CSA) and Decision-Triggered Sampling Algorithm (DTSA), each with its own transmission rule, are presented and compared against a fully-batch algorithm named Batch Sampling Algorithm (BSA). The CSA operates as a time-aware detector by incorporating the time of each transmission in the detection rule. The proposed framework encompasses the gas dispersion model into the FC's decision rule and leverages real-time weather measurements. The case study involves an accidental dispersion of carbon dioxide (CO2). System performances are evaluated in terms of the receiver operating characteristic (ROC) curve as well as average decision delay and communication cost.
... Our approach demonstrates superior performance in terms of accuracy (99.49%) and precision (100%). Among the comparable models, the one proposed by Yin et al. [26] exhibits competitive performance, with comparable accuracy and F1-score. Additionally, Nguyen et al. [27] introduced a model with a higher F1-score (96.98%), albeit at a slightly lower accuracy when compared to our method. ...
People spend most of their time indoors by choice or by need. Carbon dioxide (CO2) accumulation can cause various adverse health effects, including vertigo, headache, and fatigue. Therefore, monitoring indoor air quality(IAQ) is necessary for various health reasons. The market is flooded with air quality monitoring devices. However, the ordinary public does not make use of them because they are expensive and difficult to obtain. Several research studies have been carried out to monitor indoor air quality with the help of the Internet of Things(IoT), which has greatly simplified the method for monitoring IAQ. In this research, we offer an improved IoT based IAQ monitoring system with AI-powered recommendations. Our suggested system relies on the Message Queuing Telemetry Transport(MQTT) protocol for communication between IoT devices. In addition, the gathered CO2 occupancy data is used together with the deep learning approach of Long Short-Term Memory and Autoencoder (LSTM-AE) to detect anomalies or outliers in CO2 concentrations. Due to a close connection between air quality and human health and well-being, the detection of anomalies in the data of IAQ has emerged as an essential topic of study. Anomalies requiring the observation of correlations spanning numerous data points (i.e., often referred to as long-term dependencies) were not detectable by conventional statistical and basic machine learning (ML) related techniques in the sector of IAQ. Hence this research uses the LSTM-AE model to address this issue. In comparison to previous similar models, our experimental results on a generated CO2 occupancy time series reveal a robust and powerful accuracy of 99.49%.
... The pixel non-uniformity (PNU) metro system is mentioned as the subway or metro transportation network operated by PNU inside its surrounding regions or campus [4]. The metro system is designed to offer proficient, supportable, and suitable transportation ...
Climate control in a pixel non-uniformity metro system includes regulating the air, humidity, and temperature quality within metro trains and stations to ensure passenger comfort and safety. The climate control system in a PNU metro system combines intelligent algorithms, energy-efficient practices, and advanced technologies to make a healthy and comfortable environment for passengers while reducing energy consumption. The proposed an automated climate control using an improved salp swarm algorithm with an optimal ensemble learning technique examines the underlying factors, including indoor air temperature, wind direction, indoor air relative humidity, light sensor 1 (wavelength), return air relative humidity, supply air temperature, wind speed, supply air relative humidity, airflow rate, and return air temperature. Moreover, this new proposed technique applies ISSA to elect an optimal set of features. Then, the climate control process takes place using an ensemble learning approach comprising long short-term memory, gated recurrent unit, and recurrent neural network. Lastly, the Harris hawks optimization algorithm can be employed to adjust the hyperparameters related to the ensemble learning models. The extensive results demonstrated the supremacy of the proposed algorithms over other approaches to the climate control process on PNU metro systems.
