A high-sensitive wireless and passive surface acoustic wave (SAW) humidity sensor operating at 433 MHz was designed and developed. SiO ₂ thin-film with large specific surface was employed as the humidity sensitive interface and deposited along the sensing area of the SAW device patterned by reflective delay line by using plasma enhanced chemical vapor deposition (PECVD) to build the sensing chip. Coupling of modes (COM) model was utilized to simulate the SAW device prior to device fabrication to achieve low time-domain insertion and high signal-to-noise ratio (SNR). The transceiver based on frequency-stepped continuous wave (FSCW) and planar dipole antennas were developed to construct successfully the wireless measurement system. In wireless humidity sensing experiments, high humidity sensitivity of 11.8 °/% RH in measurement range of 45%-90% RH was achieved at the wireless distance of 0.5 m. The corresponding response and recovery time were evaluated as 60 s and 90 s respectively. Also, excellent repeatability and stability in humidity sensing were obtained.

Electronic nose is a kind of widely-used artificial olfactory system for the detection and classification of volatile organic compounds. The high dimensionality of data collected by electronic noses can hinder the process of pattern recognition. Thus, the feature selection is an essential stage in building a robust and accurate model for gas recognition. This paper proposed an improved grey wolf optimizer (GWO) based algorithm for feature selection and applied it on electronic nose data for the first time. Two mechanisms are employed for the proposed algorithm. The first mechanism contains two novel binary transform approaches, which are used for searching feature subset from electronic nose data that maximizing the classification accuracy while minimizing the number of features. The second mechanism is based on the adaptive restart approach, which attempts to further enhance the search capability and stability of the algorithm. The proposed algorithm is compared with five efficient feature selection algorithms on three electronic nose data sets. Three classifiers and multiple assessment indicators are used to evaluate the performance of algorithm. The experimental results show that the proposed algorithm can effectively select the feature subsets that are conducive to gas recognition, which can improve the performance of the electronic nose.

Baseline drift caused by slowly changing environment and other instability factors affects significantly the performance of gas sensors, resulting in reduced accuracy of gas classification and quantification of the electronic nose. In this work, a two-stage method is proposed for real-time sensor baseline drift compensation based on estimation theory and piecewise linear approximation. In the first stage, the linear information from the baseline before exposure is extracted for prediction. The second stage continuously predicts changing linear parameters during exposure by combining temperature change information and time series information, and then the baseline drift is compensated by subtracting the predicted baseline from the real sensor response. The proposed method is compared to three efficient algorithms and the experiments are conducted towards two simulated datasets and two surface acoustic wave sensor datasets. The experimental results prove the effectiveness of the proposed algorithm. Moreover, the proposed method can recover the true response signal under different ambient temperatures in real-time, which can guide the future design of low-power and low-cost rapid detection systems.

Deep learning has demonstrated great potential in the field of gas recognition using an electronic nose. However, owing to the limited resources of embedded devices, they can only meet the computing and storage requirements of lightweight deep neural network (DNN) models, which reduces the classification accuracy. This study proposes three curriculum learning (CL) approaches to improve the performance of lightweight DNN models for end-to-end gas recognition, including a domain knowledge-based approach and two approaches that automatically arrange the curriculum. Among the two automated methods, the first uses the correlation strength between sample responses to measure the difficulty of the training samples, and the second is based on the classification difficulty and disagreements among multiple teacher models. Furthermore, our approaches are combined with a mixup-based data augmentation method to apply CL to small datasets. The superiority of the proposed approaches was verified using 14 electronic nose datasets. The experimental results demonstrate that these approaches can be used on different types of datasets while boosting the gas recognition accuracy and $F1$ score of various DNN models by 9.9% and 10.9% on average, respectively, which can guide the future design of high-accuracy gas sensing embedded systems.

Surface acoustic wave (SAW) gas sensors are widely used in electronic nose systems due to high sensitivity, rapid response, and wide detection range. However, their baseline is highly susceptible to significant drift caused by environmental temperature/humidity fluctuations, severely degrading gas concentration detection accuracy. To address this, this paper proposes a baseline drift prediction method for SAW sensors, termed the Dynamic Recurrent Kalman Network (DRKN). DRKN employs a three-stage prediction-correction-denoising architecture. It integrates ambient temperature/humidity variations, combines the dynamic modeling capability of RNNs with the state estimation framework of KalmanNet, and incorporates a segment-optimized denoising mechanism to achieve precise prediction of nonlinear baseline drift. Experimental results on two real-world collected datasets demonstrate the superior performance of DRKN in both short- and long-term prediction tasks. Notably, for long-term prediction, DRKN outperforms the suboptimal model by 90% and 58% on the respective datasets. This approach provides an effective solution for mitigating baseline drift and enhancing SAW gas sensor accuracy.