Search

This paper introduces a predictive combustion-engine controller for efficient fulfillment of constraints on peak pressure, exhaust temperature and NO emissions. The controller utilizes a in-cylinder pressure sensor, actuation of multiple fuel injections and numerical optimization. A physical model is used to predict pressure changes due to fuel-injection variation. The predicted pressure is used to compute changes in fuel-injection timings and durations by solving a model-predictive control problem every engine cycle. A method for separating the heat-release among different injections is presented along with a method for varying the number of injections, since the optimal number of injections varies with operating conditions. Experimental results are presented showing controller performance in the mid-load region of a Scania D13 heavy-duty engine, running on a gasoline.

Partially premixed combustion (PPC) is a hybrid combustion of Homogeneous Charge Compression Ignition (HCCI) and Diesel Combustion (DC), which has a great potential in reducing the fuel consumption. PPC has a longer premixed time than DC but lesser than HCCI. Therefore it is more stratified than HCCI. This paper first presents results on how the stratification using multiple injections and different EGR influence the efficiency and proposes a control framework for PPC stratification control inspired from the experiments. The control framework is validated in transient operations. Results in both steady and transient operations demonstrated that the more stratified PPC with multiple injections has a lower fuel consumption.

Gasoline-fueled Partially Premixed Combustion is an advanced combustion concept approaching high efficiency as well as low emissions. The most challenging task on controlling a gasoline PPC engine is to regulate the maximum pressure rise rate to reduce engine noise and durability problem. A trade-off relationship between pressure rise rate and soot emissions is observed as a function of pilot injection event. In this paper, a control-oriented model is developed to predict in-cylinder pressure and engine outputs through fuel injection events. Thereafter, two controllers structured with PI and MPC are designed and evaluated separately. Simulation results show that, both controllers satisfy the objective of achieving desired engine load and combustion phasing, with the constraints of pressure rise rate and soot emissions simultaneously. MPC controller produces a smoother transient move with less overshoot, comparing with PI controller with a fast response.

Advanced combustion concepts like partially premixed combustion (PPC) are sensitive to inlet gas conditions especially inlet temperature. To estimate the combustion parameters like ignition delay on a cycle-to-cycle basis, both complementary filter and extended Kalman filter were designed to estimate the inlet temperature, and the ignition delay was calculated with properly estimated temperature. A two valve fast thermal management system was used to control the inlet temperature, and a PI controller with the lambda tuning method was used. The estimation and control algorithm were verified on a six-cylinder heavy duty engine.

The ignition delay is an important quantity in low temperature combustion concepts where a prolonged ignition delay gives an enhanced fuel-air mixing process, favourable for decreased particulate and NOxemission levels. This article experimentally evaluates three different low-order physics-based correlation models and their ability to predict the ignition delay for the purpose of model-based controller design. The models relate the state of the gas mixture after the point of fuel injection in order to predict the ignition delay. The experiments were all performed on a six-cylinder direct-injection Scania diesel engine with a fuel mixture of 80 vol% gasoline and 20 vol% n-heptane. The results showed that ignition-delay variation was not easily predicted when the injection timing was varied close to top-dead-centre but that the model performance satisfactorily during intake-condition variations. The results also showed that a more complex model did not necessarily give an increased prediction performance.

Partially Premixed combustion (PPC) is a promising combustion concept to achieve high engine efficiency. The combustion-timing of PPC is affected by both inlet-charge condition and the fuel injection, a simple map-based feed-forward control method is not sufficient for controlling the combustion during transient operation. This paper investigates a model-based control method to solve this control problem. A non-linear model was developed to capture the main trend of inlet-gas condition and injection. A Model Predictive Controller (MPC) was designed to control injection timing and the gas-exchange system. The controller was verified in a multi-cylinder heavy-duty PPC engine.

Partially Premixed Combustion has shown to be a promising internal combustion concept in terms of engine efficiency and emission levels, however, since the combustion process is partly kinetically controlled, unacceptably high pressure-rise rates could occur at various operating points. A remedy to this problem is the use of a pilot-fuel injection. This paper characterizes double injection effects on combustion through experimental results and presents a double injection closed-loop model predictive control strategy where the combustion phasing is controlled arbitrarily as the pressure rise rate is controlled below a upper bound. Experimental results show promising response times during set-point changes as well as load and speed disturbances.