This paper proposes a novel multidomain features-based genetic algorithm (GA) optimized artificial immune system (AIS) framework for fault detection in real systems. Different from native real-valued negative selection algorithm (RNSA) that operates in original data space, this algorithm utilizes feature space transformation and diversity factor-based GA for optimized detector distribution in nonself feature space. The proposed framework comprises three stages namely; feature extraction, unsupervised feature selection, and GA optimized AIS. In the first stage, signal processing methods are applied to extract multidomain features (time-domain statistical, frequency domain statistical, and special features) of the system. In the second stage, two unsupervised methods namely, k-NN clustering and pretraining using deep learning neural network are proposed for dominant fault-characterizing feature selection. Finally, in the third stage, the fault-characterizing feature vectors are used for system status categorization (i.e., normal, fault) using selected (fault-characterizing) features-based AIS method. The efficacy of the proposed framework is verified through experiments on motor bearing fault detection using vibration signal. The major accomplishment of the proposed combination of space transformation, feature selection and AIS (anomaly classification) techniques is the alleviation of computational burden on RNSA implementation. Moreover, GA optimized AIS fault diagnosis based on well-established features gives improved detection performance.

Agrochemicals, which are very efficacious in protecting crops, also cause environmental pollution and pose serious threats to farmers’ health upon exposure. In order to cut down the environmental and human health risks associated with agrochemical application, there is a need to develop intelligent application equipment that could detect and recognize crops/weeds, and spray precise doses of agrochemical at the right place and right time. This paper presents a machine-learning based crop/weed detection system for a tractor-mounted boom sprayer that could perform site-specific spraying on tobacco crop in fields. An SVM classifier with a carefully chosen feature combination (texture, shape, and color) for tobacco plant has been proposed and 96% classification accuracy has been achieved. The algorithm has been trained and tested on a real dataset collected in local fields with diverse changes in scale, orientation, background clutter, outdoor lighting conditions, and variation between tobacco and weeds. Performance comparison of the proposed algorithm has been made with a deep learning based classifier (customized for real-time inference). Both algorithms have been deployed on a tractor-mounted boom sprayer in tobacco fields and it has been concluded that the SVM classifier performs well in terms of accuracy (96%) and real-time inference (6 FPS) on an embedded device (Raspberry Pi 4). In comparison, the customized deep learning-based classifier has an accuracy of 100% but performs much slower (0.22 FPS) on the Raspberry Pi 4.

In this paper, a model-based fault detection and isolation (FDI) method is presented using an adaptive system identification approach. The proposed FDI method consists of three essential steps: adaptive modeling and residual generation, fault detection using adaptive hybrid threshold, and fault identification using fault severity indices. The primary task is based on current signal modeling using an input-output identification method. The modeled signal is utilized for residual generation and a dynamic and hybrid thresholding method is used for residual analysis and fault detection. Moreover, the concept of fault severity indices is incorporated for the identification of fault type and severity level. In this study, the proposed method is experimentally investigated using an induction motor testbed. Fault detection and identification is performed for broken rotor bar, inner-race and outer-race bearing faults. Experimental results are included to demonstrate the feasibility of the proposed method for fault detection and isolation. The results demonstrate robust fault detection and accurate fault isolation. The proposed fault diagnosis method provides an efficient flexible solution for improving system reliability and safety.

Selective agrochemical spraying is a highly intricate task in precision agriculture. It requires spraying equipment to distinguish between crop (plants) and weeds and perform spray operations in real-time accordingly. The study presented in this paper entails the development of two convolutional neural networks (CNNs)-based vision frameworks, i.e., Faster R-CNN and YOLOv5, for the detection and classification of tobacco crops/weeds in real time. An essential requirement for CNN is to pre-train it well on a large dataset to distinguish crops from weeds, lately the same trained network can be utilized in real fields. We present an open access image dataset (TobSet) of tobacco plants and weeds acquired from local fields at different growth stages and varying lighting conditions. The TobSet comprises 7000 images of tobacco plants and 1000 images of weeds and bare soil, taken manually with digital cameras periodically over two months. Both vision frameworks are trained and then tested using this dataset. The Faster R-CNN-based vision framework manifested supremacy over the YOLOv5-based vision framework in terms of accuracy and robustness, whereas the YOLOv5-based vision framework demonstrated faster inference. Experimental evaluation of the system is performed in tobacco fields via a four-wheeled mobile robot sprayer controlled using a computer equipped with NVIDIA GTX 1650 GPU. The results demonstrate that Faster R-CNN and YOLOv5-based vision systems can analyze plants at 10 and 16 frames per second (fps) with a classification accuracy of 98% and 94%, respectively. Moreover, the precise smart application of pesticides with the proposed system offered a 52% reduction in pesticide usage by spotting the targets only, i.e., tobacco plants.

In this paper, the challenging task of the development of a generic fault detection (FD) method is addressed. While the past FD research has primarily focused on modeling and signal-processing methods that are problem specific and require complete knowledge of the system model and fault types, this paper presents a novel layered and real-valued negative-selection algorithm (LRNSA)-based FD method independent of prior knowledge of fault types and patterns. Specifically, in the training phase, the nonself-space is divided into different layers for effective generation and distribution of detectors using normal (self) data. The major accomplishments of the proposed method are improved nonself-space coverage of the uncovered gaps (holes), followed by the formation of cluster detector with large radius. To test the capabilities of the developed method, the generated specialized detector distribution is studied using bearing fault modeling in a three-phase induction motor. The proposed method is subsequently investigated and validated by applying it to an actual induction motor for different types of bearing faults. Finally, the comparative results on the benchmark dataset demonstrate the superiority of the proposed method compared to the state-of-the-art machine learning algorithms in terms of higher FD accuracy and quick detection with reduced online detection time.