Logical control scheme with real-time statistical learning for residual gas fraction in IC engines

Dual fuel operation in compression ignition engines is an effective way to reduce the NOx emission. Within the certain range of fuel premixing ratio (PR), the dual fuel operation in CI-engines results in improved thermal efficiency. The dual fuel CI-engine has relatively higher cyclic variations in comparison to conventional CI-engine which limits the range of fuel premixing in dual fuel CI-engine. The cyclic variations in air-fuel ratio, cylinder wall temperature, and residual gas fraction are the major factors, which governs the variations in combustion parameters. The cyclic variations in combustion need to control for stable engine operation. The present study estimates the cyclic air-fuel ratio, cylinder wall temperature, and residual gas fraction from the measured in-cylinder pressure data of dual-fuel CI-engine. The experiments are performed on a modified single cylinder CI-engine equipped with a separate port fuel injector and its controller to operate an engine in dual fuel mode. In this study, 1500 consecutive engine cycles are recorded, and air-fuel ratio, cylinder wall temperature, and the residual gas fraction is estimated for each cycle. Pressure moment method is used to estimate the cyclic air-fuel ratio. The cyclic cylinder wall temperature is calculated by determining the inversion angle. The cyclic variations are analyzed using statistical methods.

In this paper, a stochastic optimal control scheme for the air-fuel ratio is proposed, which considers the cyclic variations of the residual gas fraction (RGF). Initially, a cylinder pressure-based measurement of the RGF is derived by following the physics of inlet-exhaust process. Then, a dynamical model is presented to describe the cyclic variation of the air charge, fuel charge, and combustion products under a cyclically varied RGF, where the RGF is modeled as a Markovian stochastic process. Using this model, a feedback control law is derived, which optimizes the quadratic cost function in the stochastic sense with respect to the stochastic property of the residual gas. The cost function reflects the tradeoff between the accuracy of the regulation of the air-fuel ratio with the fluctuation in the fuel injection. Finally, a sampling process-based statistical analysis for the RGF is presented based on the experiments conducted on a full-scaled gasoline engine test bench, and the proposed control law is validated based on a numerical simulation and experiments.

The in-cylinder RGF (residual gas fraction) of internal combustion engines for new combustion concepts, such as CAI (controlled auto ignition) or HCCI (homogenous charged compression ignition), is a major parameter that affects the combustion characteristics. Thus, measurement or prediction of the cycle-by-cycle RGF and investigation into the relation between the RGF and the combustion phenomena are critical issues. However, on-line prediction of the cycle-by-cycle RGF during engine testing is not always practical due to the requirement of expensive, fast response exhaust-gas analyzers and/or theoretical models that are just too slow for application. In this study, an on-line model that can predict the RGF of each engine cycle and cylinder during the experiment in the test cell has been developed. This enhanced model can predict the in-cylinder charge conditions of each engine cycle during the test in three seconds by using the measured dynamic pressures of the intake, exhaust, and cylinder as the boundary conditions. A Fortran77 code was generated to solve the 1-D MOC (method of characteristics). This code was linked to Labview DAQ as a form of DLL (dynamic link library) to obtain three boundary pressures for each cycle. The model was verified at various speeds and valve timings under the CAI mode by comparing the results with those of the commercial code, GT-Power.

<div class="htmlview paragraph">The paper describes a novel method for calculating the residual gas fraction and the trapping efficiency in a 2 stroke engine. Assuming one dimensional compressible flow through the inlet and exhaust ports, the method estimates the instantaneous mass flowing in and out from the combustion chamber; later the residual gas fraction and trapping efficiency are estimated combining together the perfect displacement and perfect mixing scavenging models. It is assumed that when the intake port opens, the fresh mixture is pushing out the burned charge without any mixing and after a multiple of the time needed for the largest eddy to perform one rotation, the two gasses are instantly mixed up together and expelled. The result is a very simple algorithm that does not require much computational time and is able to estimate with high level of precision the trapping efficiency and the residual gas fraction in 2 stroke engines. The tuning and the validation of this algorithm are performed by measuring the trapping efficiency. These measurements are conducted with an innovative technique which consists in measuring the inlet and exhaust temperature, and calculation of the blow down temperature from the pressure trace. The model was tested using a mini high speed 2 stroke HCCI engine fuelled with diethyl ether between 10,800 and 19,000 [rpm].</div>

In this paper, the problem of stochastic L2 disturbance attenuation of the air-fuel ratio is investigated with consideration of cyclic variation of the residual gas fraction (RGF). A stochastic robust controller is designed based on a discrete-time dynamic model in which the RGF is modeled as a stochastic process with Markovian property. Finally, the sampling process-based statistical analysis for the RGF and the validation of the proposed control law are presented through the experiments conducted on a gasoline engine test bench.

<div class="htmlview paragraph">The paper presents an original review and extension of existing mathematical models for on-line residual gas fraction estimation. The resulting model has first of all been extended to take into account also the presence of externally recirculated exhaust gas (external EGR), and then critically analyzed to highlight the importance of a correct Intake Valve Opening and Exhaust Valve Closing effective position identification. As shown in the paper, such quantities may be evaluated by using experimental data, either acquired in the test-cell or on a valve flow bench.</div> <div class="htmlview paragraph">The main objective is to obtain a simple and reliable model (that could be run in real time within the engine control unit) also in presence of Variable Valve Timing (VVT, both on intake and exhaust valves) and external Exhaust Gas Recirculation (EGR) systems. In fact, the two main contributions to residual gas fraction (backflow of the burned gas during the valve overlap period, and amount of gas trapped within the cylinder) are strongly affected by intake and exhaust valves timing, and EGR flow should be taken into account in order to determine the total exhaust gas mass within the cylinder at IVC. Therefore, real time estimation of residual gas mass and composition is crucial for designing VVT and EGR management strategies that allow an optimal control of the combustion process.</div> <div class="htmlview paragraph">The new model has been applied to experimental data acquired on a 3.2 liter V6 GDI engine, equipped with intake and exhaust Variable Valve Timing systems. Tests were performed throughout the engine operating range for different combinations of intake and exhaust valve timings, while varying EGR flow. Model results are in good agreement with other measured quantities (such as Spark Advance angle and NO<sub>x</sub> emissions), and the proposed approach therefore represents a powerful tool for on-board optimal combustion control.</div>

<div class="htmlview paragraph">We propose a model of in-cylinder air mass and residual gas fraction of a turbocharged SI engine with Variable Valve Timing (VVT) actuators. VVT devices are used to produce internal exhaust gas recirculation at part load, providing beneficial effects in terms of fuel consumption and pollutant emissions. At full load, VVT actuators permit to push back knock limit by scavenging fresh air to the exhaust pipes. Modeling in-cylinder composition is an essential task for control purpose. Actually, VVT actuators affect in-cylinder fresh air charge. This has an impact on engine torque output (leading to driveability problems), and on Fuel/Air Ratio (leading to pollution peaks). In this paper, we present a model of in-cylinder air mass and residual gas fraction using only commercial-line sensors (engine speed, intake manifold pressure and VVT actuators positions). It is designed for real-time control purpose. The model does not necessitate a lot of calibration time. The proposed method consists in modeling the quantity of fresh air going through the intake valve. The model is composed of three terms. All these terms have their own physical meaning: first term gives in-cylinder total mass, second term gives the gas mass that goes through both valves during the valve overlap (either backflow of burned gas, or scavenging of fresh air), and last term models the residual gas mass staying in the cylinder from one cycle to the next one.</div>

High viscosity of quench oil is a critical problem of quench system in ethylene cracking furnaces in petrochemical plant, due to its influences on the safety and stability of equipments, meanwhile, the variety of viscosity of quench oil has a negative impact on yield of ethylene and other chemical products. This paper presents a new statistical learning model to forecast the real-time variety of quench oil viscosity based on statistical algorithm and machine learning method. Firstly, statistical algorithm is applied to reduce dimension of parameters, secondly, fitting real-time predictive model through machine learning method. The simulation results shows that this model can monitor the variety of viscosity per hour according to identified controllable parameters which are highly correlated with viscosity of quench oil.

The automotive industry has long prioritized the search for more efficient engines, driven by environmental concerns, fuel prices, and consumer demand for fuel-efficient vehicles. Various technologies and strategies are explored to enhance performance while reducing fuel consumption and emissions. Despite potential combustion stability issues, techniques like using lean mixtures or retaining exhaust gases are commonly employed. Ozone seeding has emerged as a combustion enhancer for internal combustion engines, particularly effective under lean mixtures or high residual gas levels, aiming to stabilize de-throttling conditions, minimize pumping losses, and enhance engine efficiency by enhancing the reactivity of the mixture through the release of radicals that facilitate fuel oxidation. This study investigates the impact of ozone addition on engine performance, emissions, and fuel consumption using a turbocharged SI vehicular engine on a dynamometer bench under different loads. Ozone was introduced to the engine intake using an ozone generator with compressed air, considering two mixture dilution strategies: fresh air and residual gases. Results show that ozone application, with Brazilian Gasoline Type C, induces autoignition of the end-gas under specific low-load, low-engine speed conditions. Up to 1000 ppm, ozone has no discernible impact on DISI engine operation under excess air dilution, maintaining performance equivalent to the baseline. For specific operating conditions with 3 and 4 bar BMEP, where ozone stabilizes combustion with higher residual gas fraction (RGF) percentages, there is a notable reduction in specific fuel consumption, suggesting a potential 1%–1.5% increase in brake efficiency. However, ozone seeding with high RGF percentages leads to a slight increase in CO emissions, elevated NOX emissions, and a tendency to reduce THC emissions up to a certain RGF limit.

Skip-firing (or cylinder de-activation) was assessed as a method of sampling CO2 from directly in the cylinder at higher speeds than previously possible. CO2 was directly sampled from one cylinder of a 1-litre 3-cylinder gasoline engine to determine the residual gas fraction using a fast response CO/CO2 analyser. Acquisition of data for similar measurements is typically limited to engine speeds of below 1300 rpm to allow full resolution of the sample, through the analyser that has an 8 millisecond finite response time. In order to sample in-cylinder CO2 at higher engine speeds a skip-firing method is developed. By shutting off ignition intermittently during engine operation, the residual CO2 from the last firing cycle can be measured at significantly higher engine speeds. Comparison of residual gas fraction CO2 at low speeds for normal and skip-fire operation shows good correlation. This suggests that skip-firing is a suitable method for directly measuring internal exhaust gas recirculation up to at least 3000 rpm. The measurements obtained may provide a useful tool for validating internal exhaust gas recirculation models and could be used to calculate combustion air-fuel ratio from the CO and CO2 content of the burned gas. These are typically complicated parameters to predict due to the slow response time and sensitivity to hydrocarbons of wide-band oxygen sensors. A differing pattern of residual gas fraction change with increasing speed was seen between normal and skipfire operation.

In this article, we propose an adaptive robust nonlinear optimal sliding mode control (SMC) using the optimal homotopy asymptotic method (RNOSC‐OHAM) for maximizing wind power capture. Because of the unstable nature of the wind and the presence of uncertainties and disturbances in the structure of the wind turbine, the optimal controller cannot provide robustness against uncertainties. Hence, we integrated this approach with a proportional–integral–derivative SMC (PID‐SMC) scheme to ensure the robustness of the system. The presented hybrid control input signal involves two components: a nonlinear optimal control law that provides the optimal performance of the nominal system and an adaptive PID‐SMC law. The adaptive PID‐SMC law employs a PID sliding surface along with adaptive mechanisms to create the robustness of the nonlinear wind turbine system. The nonlinear optimal control policy is designed by addressing the partial differential Hamilton–Jacobi–Bellman (HJB) equation. This equation is approximated using the OHAM. This combination strategy reduces the final time of the optimal control problems, provides desired responses and improves performance. Additionally, the RNOSC‐OHAM controller facilitates the safe performance of the wind turbine under uncertainties and maximizes the wind power captured. To evaluate the performance of RNOSC‐OHAM, the results of the presented algorithm are compared with some existing control schemes. The results indicate that the designed RNOSC‐OHAM controller is very rapid and needs few iterations and computational costs. Indeed, the present control scheme exhibits the best characteristics and has the fastest transient response. The RNOSC‐OHAM controller offers a proper balance between enhancing aerodynamic power capture and minimizing low‐speed shaft fluctuations using a small control input.

There is growing momentum to develop statistical learning (SL) methods as an alternative to conventional genome-wide association studies (GWAS). Methods such as random forests (RF) and gradient boosting machine (GBM) result in variable importance measures that indicate how well each single-nucleotide polymorphism (SNP) predicts the phenotype. For RF, it has been shown that variable importance measures are systematically affected by minor allele frequency (MAF) and linkage disequilibrium (LD). To establish RF and GBM as viable alternatives for analyzing genome-wide data, it is necessary to address this potential bias and show that SL methods do not significantly under-perform conventional GWAS methods. Both LD and MAF have a significant impact on the variable importance measures commonly used in RF and GBM. Dividing SNPs into overlapping subsets with approximate linkage equilibrium and applying SL methods to each subset successfully reduces the impact of LD. A welcome side effect of this approach is a dramatic reduction in parallel computing time, increasing the feasibility of applying SL methods to large datasets. The created subsets also facilitate a potential correction for the effect of MAF using pseudocovariates. Simulations using simulated SNPs embedded in empirical data-assessing varying effect sizes, minor allele frequencies and LD patterns-suggest that the sensitivity to detect effects is often improved by subsetting and does not significantly under-perform the Armitage trend test, even under ideal conditions for the trend test. Code for the LD subsetting algorithm and pseudocovariate correction is available at http://www.nd.edu/~glubke/code.html.

<div class="section abstract"><div class="htmlview paragraph">An accurate prediction of residual burned gas within the combustion chamber is important to quantify for development of modern engines, especially so for those with internally recycled burned gases and HCCI operations. A wall-guided GDI engine has been fitted with an in-cylinder sampling probe attached to a fast response NDIR analyser to measure in-situ the cycle-by-cycle trapped residual gas. The results have been compared with a model which predicts the trapped residual gas fraction based on heat release rate calculated from the cylinder pressure data and other factors. The inlet and exhaust valve timings were varied to produce a range of Residual Gas Fraction (RGF) conditions and the results were compared between the actual measured CO2 values and those predicted by the model, which shows that the RGF value derived from the exhaust gas temperature and pressure measurement at EVC is consistently overestimated by 5% over those based on the CO2 concentrations. The implications for engines with internally recycled burned gases and HCCI combustion development were considered. It is shown that there is an optimum valve overlap period for a given operating condition that produces highest indicated thermal efficiency and optimised combustion phasing. In addition, the minimum residual gas concentration should be employed for the least pumping loss during the recompression process when the negative valve overlap method is used to achieve CAI/HCCI combustion for a given engine output.</div></div>

<div class="htmlview paragraph">The influence of residual gases on the cycle-to-cycle variations in engine combustion was investigated. Two-photon planar laser-induced fluorescence was used for 2D-visualization of residual gas water. In order to avoid influence from fuel fluorescence and inhomogeneities premixed natural gas was used as fuel. Measurements were conducted at different load conditions with varying inlet manifold pressure. To find out how the residual gas distribution influences the combustion process the pressure development during combustion was monitored. From the pressure information a measure of the combustion rate at different phases of the flame development was calculated. The correlation between residual gas distribution and combustion rate was evaluated on a cycle to cycle basis. The results show that with an inlet manifold pressure of 0.3 bar the correlation between residual gas fraction and rate of combustion were 0.5-0.6. At full load though, lower correlation was found.</div>

A hybrid spark ignition/homogeneous charge compression ignition engine model for simulating the spark ignition to homogeneous charge compression ignition transition in a four-cylinder gasoline engine is presented. This hybrid spark ignition/homogeneous charge compression ignition model has been developed to provide help in understanding the complicated spark ignition to homogeneous charge compression ignition transition process and in tackling the control problems associated with it. The model was built in Matlab/Simulink, coupled with a tabulated stochastic reactor model and a detailed chemical kinetics mechanism. Typical computational times encountered in the simulation using this model were of the order of seconds for one engine cycle. The model was validated in steady-state spark ignition and homogeneous charge compression ignition operation by experimental data, followed by model-based parametric studies in homogeneous charge compression ignition operation. Then, a spark ignition to homogeneous charge compression ignition mode transition was simulated using the model. The results showed good agreement with experimental data, in terms of cylinder pressure, combustion phasing, air/fuel ratio, residual gas fraction and nitrogen oxides. The high computational efficiency and relatively high accuracy of this model enable its good applicability in experimental design and controller development, and potential use in hardware-in-loop applications.