This article considers the application and refinement of artificial neural network methods for the prediction of NO x emissions from a high-speed direct injection diesel engine over a wide range of engine operating conditions. The relative computational cost and performance of two backpropagation algorithms, Levenberg–Marquardt and Bayesian regularization, for this application are compared, with the Levenberg–Marquardt algorithm demonstrating a significant cost advantage. This work also assesses the performance of two alternative filtering approaches, a p-value test and the Pearson correlation coefficient, for reducing the required number of input variables to the model. The p-value test identified 32 input parameters of significance, whereas the Pearson correlation test highlighted 14 significant parameters while additionally providing a ranking of their relative importance. Finally, the article compares the predictive performance of the models generated by the two filtering methods. Overall, both models show good agreement to the experimental data with the model created using the Pearson correlation test showing improved performance in the low-NO x region.

The effects of different exhaust gas recirculation (EGR) strategies on engine efficiency and the resulting energy flows at two speed/load conditions (1500 r/min/6.8 bar net indicated mean effective pressure (nIMEP) and 1750 r/min/13.5 bar nIMEP) were studied using a first law analysis approach. The EGR strategies tested were as follows: cooled high-pressure exhaust gas recirculation (baseline), the application of exhaust gas recirculation with the swirl flap closed and the use of exhaust gas recirculation under constant λ conditions. The closed swirl flap exhaust gas recirculation strategy reduced brake efficiency under high load conditions and increased heat transfer to the coolant for both load cases. Soot and CO emissions increased at high load, however, with an increase in NOx relative to the baseline case. The constant λ exhaust gas recirculation strategy reduced brake efficiency under low load, as well as the heat flow to the coolant for both load cases. The constant λ exhaust gas recirculation strategy benefited smoke emissions and increased combustion exhaust gas recirculation tolerance, albeit with a penalty in NOx emission.

Accurate measurement of exhaust gas temperature in internal combustion engines is essential for a wide variety of monitoring and design purposes. Typically these measurements are made with thermocouples, which may vary in size from 0.05 mm (for fast response applications) to a few millimetres. In this work, the exhaust of a single cylinder diesel engine has been instrumented both with a fast-response probe (comprising of a 50.8 μm, 127 μm and a 254 μm thermocouple) and a standard 3 mm sheathed thermocouple in order to assess the performance of these sensors at two speed/load conditions. The experimental results show that the measured timeaverage exhaust temperature is dependent on the sensor size, with the smaller thermocouples indicating a lower average temperature for both speed/load conditions. Subject to operating conditions, measurement discrepancies of up to ~80 K have been observed between the different thermocouples used. Thermocouple modelling supports the experimental trends and shows that the effect of conduction is inversely proportional to the thermocouple junction size-an effect attributed to changes in the thermal inertia of the device. This conduction error is not typically considered in the literature for exhaust gas temperature measurement. Modelling results also show that radiative heat transfer is small compared to the effect of conduction on the measurements. Finally, a new dynamic response thermocouple compensation method is presented, in order to correct for the dynamic error induced by the thermocouples. This technique recovers the "true" gas temperature with a maximum error of ~1.5-2 % in peak temperature depending on speed/load conditions.

Abstract. Low-cost air quality sensors offer significant potential for enhancing urban air quality networks by providing higher-spatiotemporal-resolution data needed, for example, for evaluation of air quality interventions. However, these sensors present methodological and deployment challenges which have historically limited operational ability. These include variability in performance characteristics and sensitivity to environmental conditions. In this work, we investigate field “baselining” and interference correction using random forest regression methods for low-cost sensing of NO2, PM10 (particulate matter) and PM2.5. Model performance is explored using data obtained over a 7-month period by real-world field sensor deployment alongside reference method instrumentation. Workflows and processes developed are shown to be effective in normalising variable sensor baseline offsets and reducing uncertainty in sensor response arising from environmental interferences. We demonstrate improvements of between 37 % and 94 % in the mean absolute error term of fully corrected sensor datasets; this is equivalent to performance within ±2.6 ppb of the reference method for NO2, ±4.4 µg m−3 for PM10 and ±2.7 µg m−3 for PM2.5. Expanded-uncertainty estimates for PM10 and PM2.5 correction models are shown to meet performance criteria recommended by European air quality legislation, whilst that of the NO2 correction model was found to be narrowly (∼5 %) outside of its acceptance envelope. Expanded-uncertainty estimates for corrected sensor datasets not used in model training were 29 %, 21 % and 27 % for NO2, PM10 and PM2.5 respectively.

The understanding and prediction of NOx emissions formation mechanisms during engine transients are critical to the monitoring of real driving emissions. While many studies focus on the engine out NOx formation and treatment, few studies consider cyclic transient NOx emissions due to the low time resolution of conventional emission analysers. Increased computational power and substantial quantities of accessible engine testing data have made ANN a suitable tool for the prediction of transient NOx emissions. In this study, the transient predictive ability of artificial neural networks where a large number of engine testing data are available has been studied extensively. Significantly, the proposed transient model is trained from steady-state engine testing data. The trained data with 14 input features are provided with transient signals which are available from most engine testing facilities. With the help of a state-of-art high-speed NOx analyser, the predicted transient NOx emissions are compared with crank-angle resolved NOx measurements taken from a high-speed light duty diesel engine at test conditions both with and without EGR. The results show that the ANN model is capable of predicting transient NOx emissions without training from crank-angle resolved data. Significant differences are captured between the predicted transient and the slow-response NOx emissions (which are consistent with the cycle-resolved transient emissions measurements). A particular strength is found for increasing load steps where the instantaneous NOx emissions predicted by the ANN model are well matched to the fast-NOx analyser measurements. The results of this work indicate that ANN modelling could strongly contribute to the understanding of real driving emissions.

Chromium-molybdenum alloy steel pistons, which have been used in commercial vehicle applications for some time, have more recently been proposed as a means of improving thermal efficiency in lightduty applications. This work reports a comparison of the effects of geometrically similar aluminium and steel pistons on the combustion characteristics and energy flows on a single cylinder high-speed direct injection diesel research engine tested at two speed / load conditions (1500 rpm / 6.9 bar nIMEP and 2000 rpm / 25.8 bar nIMEP) both with and without EGR. The results indicate that changing to an alloy steel piston can provide a significant benefit in brake thermal efficiency at part-load and a reduced (but nonnegligible) benefit at the high-load condition and also a reduction in fuel consumption. These benefits were attributed primarily to a reduction in friction losses. In terms of energy transfer, switching to the steel piston design was shown to reduce heat transfer to the coolant, consistent with lower friction work and reduced conduction through the ring pack, and increase the energy transfer to the oil. Piston blowby was also greatly reduced. Ignition delay times and overall combustion durations were reduced with the steel piston design, possibly indicative of higher piston surface temperatures.

This study looked into the application of active thermal coatings on the surfaces of the combustion chamber as a method of improving the thermal efficiency of internal combustion engines. The active thermal coating was applied to a production aluminium piston and its performance was compared against a reference aluminium piston on a single-cylinder diesel engine. The two pistons were tested over a wide range of speed/load conditions and the effects of EGR and combustion phasing on engine performance and tailpipe emissions were also investigated. A detailed energy balance approach was employed to study the thermal behaviour of the active thermal coating. In general, improvements in indicated specific fuel consumption were not statistically significant for the coated piston over the whole test matrix. Mean exhaust temperature showed a marginal increase with the coated piston of up to 6 °C. However, the normalised exhaust enthalpy showed a reduction (apart from the higher speed/load conditions when no EGR was applied). Energy transfer to the coolant was reduced by as much as 1.5 percentage points, in agreement with the expected reduction in piston heat transfer, across all operating conditions. Finally, soot emissions were increased with the coated piston, with the biggest differences between the coated and non-coated pistons observed at the lower speed/load conditions.

The predictive ability of artificial neural networks where a large number of experimental data are available, has been studied extensively. Studies have shown that ANN models are capable of accurately predicting NO x emissions from engines under various operating conditions and different fuel types when trained well. One of the major advantages of an ANN model is its ability to relearn when new experimental data is available, thus continuously improving its accuracy. The present work explored the potential of an ANN model to predict NO x emissions for various engine configurations outside its training envelop. This work also looked into quantifying the amount of new data required to improve the accuracy of the model when exposed to unknown conditions. The chosen ANN model was constructed using data from a high-speed direct injection diesel engine and is capable of accurate NO x emissions over a wide range of operating conditions. The optimized network utilized 14 input parameters and is using 6 neurons in a single hidden layer feed-forward neural network. Experimental data from the various engine configurations tested, were then used to predict NO x from the existing ANN model. The results indicate that when the new data are within the baseline training envelop, the ANN model is capable of accurate NO x prediction even when there are substantial changes in engine configuration such as piston material. Similar results were also observed when the injector nozzle is changed. However, the model's performance drops significantly when new data, outside the baseline training envelop, were employed in- * Address all correspondence to this author.dicating that additional training is required. As such, various methods for retraining the ANN model were explored with the selected method showing the best compromise between new-data accuracy and old-data accuracy retention. The retrained ANN model developed was found to be an effective tool in predicting NO x emissions for different engine configurations and operating conditions.