Meteorological parameters such as relative humidity have a significant impact on the precision of PM2.5 measurement instruments based on light scattering. Instead of adding meteorological sensors or dehumidification devices used widely in commercial PM2.5 measurement instruments, a novel particle sensor based on multi-angle light scattering and data fusion is proposed to eliminate the effect of meteorological factors. Three photodiodes are employed to collect the scattered light flux at three distinct angles. Weather index is defined as the ratio of scattered light fluxes collected at the 40° and 55° angles, which can be used to distinguish the mass median diameter variation caused by different meteorological parameters. Simulations based on Lorenz-Mie theory and field experiments establish the feasibility of this scheme. Experimental results indicate that mass median diameter has less effect on the photodiode at the 55° angle in comparison with photodiodes at the 40° angle and 140° angle. After correction using the weather index, the photodiode at the 40° angle yielded the best results followed by photodiodes at the 55° angle and the 140° angle.

An accurate prediction of the remaining useful life (RUL) of a proton exchange membrane fuel cell (PEMFC) is of great significance for its large-scale commercialization and life extension. This paper aims to develop a PEMFC degradation prediction method that incorporates short-term and long-term predictions. In the short-term prediction, a long short-term memory (LSTM) neural network is combined with a Gaussian process regression (GPR) probabilistic model to form a hybrid LSTM-GPR model with a deep structure. The model not only can accurately forecast the nonlinear details of PEMFC degradation but also provide a reliable confidence interval for the prediction results. The results showed that the proposed LSTM-GPR model outperforms the single models in both prediction accuracy and confidence interval. For the long-term prediction, a novel RUL prediction model based on an extended Kalman filter (EKF) and GPR is proposed. The GPR model is used to solve the problem that the EKF cannot update the model parameters in the prediction stage. The results showed that the proposed EKF-GPR model can achieve better RUL prediction than the model-based approach and the data-driven approach.

Deep displacement observation is one basic means of landslide dynamic study and early warning monitoring and a key part of engineering geological investigation. In our previous work, we proposed a novel electromagnetic induction-based deep displacement sensor (I-type) to predict deep horizontal displacement and a theoretical model called equation-based equivalent loop approach (EELA) to describe its sensing characters. However in many landslide and related geological engineering cases, both horizontal displacement and vertical displacement vary apparently and dynamically so both may require monitoring. In this study, a II-type deep displacement sensor is designed by revising our I-type sensor to simultaneously monitor the deep horizontal displacement and vertical displacement variations at different depths within a sliding mass. Meanwhile, a new theoretical modeling called the numerical integration-based equivalent loop approach (NIELA) has been proposed to quantitatively depict II-type sensors’ mutual inductance properties with respect to predicted horizontal displacements and vertical displacements. After detailed examinations and comparative studies between measured mutual inductance voltage, NIELA-based mutual inductance and EELA-based mutual inductance, NIELA has verified to be an effective and quite accurate analytic model for characterization of II-type sensors. The NIELA model is widely applicable for II-type sensors’ monitoring on all kinds of landslides and other related geohazards with satisfactory estimation accuracy and calculation efficiency.

Accurate estimation of state of charge (SOC) is the basis of battery management system. The fractional-order theory is introduced into the resistance-capacitance (RC)model of lithium battery, and adaptive genetic algorithm is used to identify the parameters of the RC model. Considering the changes of internal resistance and battery aging during battery discharge, the state of health (SOH) of the battery is estimated by unscented Kalman filter (UKF), and the values of internal resistance and maximum capacity of the battery are obtained. Finally, a novel estimation algorithm of lithium battery SOC based on SOH and fractional-order adaptive extended Kalman filter (FOAEKF) is proposed. An experimental system is set up to verify the effectiveness of the proposed algorithm. The experimental results show that, compared with the traditional SOC estimation algorithms, the proposed method has higher estimation accuracy, with the average error lower than 1% and the maximum error lower than 2%.