This paper develops a new approach to estimation of in-cylinder pressure and combustion variables for cycle-to-cycle combustion control in diesel engines. Such combustion control can lead to enhancement of engine performance and efficiency as well as prevention of combustion failures in UAV diesel engines. In-cylinder pressure and combustion variables are estimated using measurements from a non-intrusive accelerometer located on the engine block. The new estimation approach is based on separating the combustion component of the acceleration signal from the non-combustion component. A time-varying first order differential equation model relating the combustion components of acceleration and pressure is then utilized. The proposed method is evaluated using experimental data from a turbocharged high speed diesel engine. The robustness of the proposed estimation algorithm is demonstrated by evaluating it with 85 experimental data sets of many different operating conditions involving both single and double (pilot and main) injections. In-cylinder pressure and combustion variables such as cumulative heat release and crank angle of 50 percent cumulative heat release (CA50) are estimated, with CA50 being a key variable needed for closed-loop cycle-to-cycle combustion control. Experimental results show that the CA50 value is estimated accurately with an RMS error of 1.45 degrees in single injection data sets involving 40 different operating conditions and an RMS error of 3.96 degrees in double injection data sets involving 45 different operating conditions.

This paper introduces a novel methodology for evaluating human driving behavior influenced by shoe type and its impact on collision risk. While human factors, such as footwear, are recognized to affect driving safety, studies quantitively assessing the effects of shoe types on safety has been limited. To address this, we utilize an autoencoder and human-centered risk assessment algorithms to investigate human driving behavior and collision risk. Experiments were conducted in various real-world driving scenarios, involving two distinct types of shoes. The autoencoder extracts features from the driving data and enables us to analyze the effects of shoe type on driving behavior. Additionally, collision risk analysis is used to verify the validity and impact of the feature extraction results on safe driving. This study contributes to enhancing our understanding of how footwear influences driver behavior and safety. Furthermore, this methodology establishes a groundwork for future research on applying quantitative evaluations to other human factors that influence driving behavior.

Abstract Detecting internal short circuits (ISCs) in a single cell connected in parallel with others is challenging because unmeasured internal currents can obscure measurable indicators such as charge loss and voltage drop from the faulty cell. In this work, we propose a new method for detecting ISCs based on the interacting multiple model (IMM) estimation technique, which can provide a probability for the occurrence of an ISC and simultaneously estimate the short-circuit resistance, indicating the severity of the ISC. The IMM relies on dynamic electrothermal models of parallel cells (nP), both for the healthy mode and short-circuit mode. The IMM technique is combined with unscented Kalman filters (UKFs) to detect internal short circuits and estimate the short-circuit resistance across various synthetic data sets that are corrupted by Gaussian noise for different values of ISC resistance. Fifty short-circuit scenarios were simulated in which one cell in a 46 P cell group underwent an ISC during a drive cycle. The short-circuit resistances ranged from 0.5 to 100 Ω, tested at ten different states of charge (SOCs). Our simulation outputs included busbar voltage, input current, and cell temperatures, which were then corrupted by Gaussian noise. Our IMM successfully detected and estimated the ISC in all fifty cases, with temperature rise remaining below 6 °C before detection, well before the onset of thermal runaway conditions.

This paper focuses on the development of a methodology to monitor the health of an engine by detecting any failures in the fuel injectors or in-cylinder pressure sensors using an accelerometer that is non-intrusively mounted on the engine block. A multi-cylinder engine with each cylinder having its own pressure sensor and injector is considered. First, a model relating the combustion component of the measured acceleration signal to the combustion component of in-cylinder pressure is proposed. Then, gains of the model are tuned to reduce the cycle-to-cycle estimation error by analyzing cycle-to-cycle variations with respect to the combustion pressure peak and engine vibration peak. Using the developed model, cylinder combustion pressures are estimated from engine vibration signals with small cycle-to-cycle estimation errors. Subsequently, a health monitoring system that can detect faults in pressure sensors, fuel injectors, and the accelerometers is proposed based on residues obtained from the difference between estimated combustion pressure and measured pressure signals. The source of the failed component can be identified uniquely by analyzing the pattern of residues. The proposed combustion pressure estimation algorithms are validated by extensive evaluation with experimental data obtained by operating a four-cylinder compression-ignition direct-injection engine with a range of experimental data. Finally, the developed health monitoring system is evaluated with various failure scenarios involving faults in the in-cylinder pressure sensor, fuel injector, and accelerometer.

<title>Abstract</title> This paper describes clustering and investigation of human driving behavior with shoe type in urban roads using autoencoder. The analysis of driving patterns for different shoe types is important as it has been known to affect safe driving due to the braking distance change. To this end, we analyzed the effect of the type of shoes on vehicle driving by using an autoencoder to train the data acquired from actual vehicle driving with two types of shoes. With successfully trained data, the driving characteristics have been clustered on various urban driving scenarios when a preceding vehicle exists. The validity and impact of the clustering results on safe driving were verified through collision risk analysis in order to investigate the safety effects of shoe types. It has been shown from vehicle tests that the proposed clustering analysis with probabilistic risk assessment presents a clear correlation between footwear choice and driving safety, with the maximum collision probability was reduced by 23%, and the maximum collision time was improved by 0.4 seconds when driving shoes were worn.

Abstract This article develops a neuro‐adaptive observer for state and nonlinear function estimation in systems with partially modeled process dynamics. The developed adaptive observer is shown to provide exponentially stable estimation errors in which both states and nonlinear functions converge to their true values. When the neural approximator has an approximation error with respect to the true nonlinear function, the observer can be used to provide an bound on the estimation error. The article does not require assumptions on the process dynamics or output equation being linear functions of neural network weights and instead assumes a reasonable affine parameter dependence in the process dynamics. A convex problem is formulated and an equivalent polytopic observer design method is developed. Finally, a hybrid estimation system that switches between a neuro‐adaptive observer for system identification and a regular nonlinear observer for state estimation is proposed. The switched operation enables parameter estimation updates whenever adequate measurements are available. The performance of the developed adaptive observer is shown through simulations for a Van der Pol oscillator and a single link robot. In the application, no manual tuning of adaptation gains is needed and estimates of both the states and the nonlinear functions converge successfully.