Search

The high thermal efficiency and reliability of the compression-ignition engine makes it the first choice for many applications. For this to continue, a reduction of the pollutant emissions is needed. One solution is the use of machine learning (ML) and model predictive control (MPC) to minimize emissions and fuel consumption, without adding substantial computational cost to the engine controller. ML is developed in this paper for both modeling engine performance and emissions and for imitating the behaviour of an Linear Parameter Varying (LPV) MPC. Using a support vector machine-based linear parameter varying model of the engine performance and emissions, a model predictive controller is implemented for a 4.5 Cummins diesel engine. This online optimized MPC solution offers advantages in minimizing the \nox~emissions and fuel consumption compared to the baseline feedforward production controller. To reduce the computational cost of this MPC, a deep learning scheme is designed to mimic the behavior of the developed controller. The performance in reducing NOx emissions at a constant load by the imitative controller is similar to that of the online optimized MPC compared to the Cummins production controller. In addition, the imitative controller requires 50 times less computation time compared to that of the online MPC optimization., Comment: Submitted to Advances in Automotive Control - 10th AAC 2022

Machine learning (ML) and a nonlinear model predictive controller (NMPC) are used in this paper to minimize the emissions and fuel consumption of a compression ignition engine. In this work machine learning is applied in two methods. In the first application, ML is used to identify a model for implementation in model predictive control optimization problems. In the second application, ML is used as a replacement of the NMPC where the ML controller learns the optimal control action by imitating or mimicking the behavior of the model predictive controller. In this study, a deep recurrent neural network including long-short term memory (LSTM) layers are used to model the emissions and performance of an industrial 4.5 liter 4-cylinder Cummins diesel engine. This model is then used for model predictive controller implementation. Then, a deep learning scheme is deployed to clone the behavior of the developed controller. In the LSTM integration, a novel scheme is used by augmenting hidden and cell states of the network in an NMPC optimization problem. The developed LSTM-NMPC and the imitative NMPC are compared with the Cummins calibrated Engine Control Unit (ECU) model in an experimentally validated engine simulation platform. Results show a significant reduction in Nitrogen Oxides (\nox) emissions and a slight decrease in the injected fuel quantity while maintaining the same load. In addition, the imitative NMPC has a similar performance as the NMPC but with a two orders of magnitude reduction of the computation time., Comment: Submitted to Control engineering Practice (Submission date: March 9, 2022) Revised version (Submission date: June 18, 2022) Accepted on July 30, 2022

Sliding mode control (SMC) is a robust and computationally efficient model-based controller design technique for highly nonlinear systems, in the presence of model and external uncertainties. However, the implementation of the conventional continuous-time SMC on digital computers is limited, due to the imprecisions caused by data sampling and quantization, and the chattering phenomena, which results in high frequency oscillations. One effective solution to minimize the effects of data sampling and quantization imprecisions is the use of higher order sliding modes. To this end, in this paper, a new formulation of an adaptive second order discrete sliding mode control (DSMC) is presented for a general class of multi-input multi-output (MIMO) uncertain nonlinear systems. Based on a Lyapunov stability argument and by invoking the new Invariance Principle, not only the asymptotic stability of the controller is guaranteed, but also the adaptation law is derived to remove the uncertainties within the nonlinear plant dynamics. The proposed adaptive tracking controller is designed and tested in real-time for a highly nonlinear control problem in spark ignition combustion engine during transient operating conditions. The simulation and real-time processor-in-the-loop (PIL) test results show that the second order single-input single-output (SISO) DSMC can improve the tracking performances up to 90%, compared to a first order SISO DSMC under sampling and quantization imprecisions, in the presence of modeling uncertainties. Moreover, it is observed that by converting the engine SISO controllers to a MIMO structure, the overall controller performance can be enhanced by 25%, compared to the SISO second order DSMC, because of the dynamics coupling consideration within the MIMO DSMC formulation., Comment: 12 pages, 7 figures, 1 table

This paper presents a new predictive second order sliding controller (PSSC) formulation for setpoint tracking of constrained linear systems. The PSSC scheme is developed by combining the concepts of model predictive control (MPC) and second order discrete sliding mode control. In order to guarantee the feasibility of the PSSC during setpoint changes, a virtual reference variable is added to the PSSC cost function to calculate the closest admissible set point. The states of the system are then driven asymptotically to this admissible setpoint by the control action of the PSSC. The performance of the proposed PSSC is evaluated for an advanced automotive engine case study, where a high fidelity physics-based model of a reactivity controlled compression ignition (RCCI) engine is utilized to serve as the virtual test-bed for the simulations. Considering the hard physical constraints on the RCCI engine states and control inputs, simultaneous tracking of engine load and optimal combustion phasing is a challenging objective to achieve. The simulation results of testing the proposed PSSC on the high fidelity RCCI model show that the developed predictive controller is able to track desired engine load and combustion phasing setpoints, with minimum steady state error, and no overshoot. Moreover, the simulation results confirm the robust tracking performance of the PSSC during transient operations, in the presence of engine cyclic variability., Comment: 6 pages, 5 figures, 2018 American Control Conferance (ACC), June 27-29, 2018, Milwaukee, WI, USA. [Accepted in Jan. 2018]

In this paper, the dynamic couplings among different inputs and outputs of a highly nonlinear spark ignition (SI) engine control problem during the cold start phase are evaluated by using relative gain array (RGA) technique. First, based on the experimental data, a multi-input multi-output model is developed to represent the engine dynamics. Second, the coupling among different inputs and outputs of the model is evaluated by using RGA technique in both open-loop and closed-loop structures. The results show that although there is an internal coupling within the engine dynamics in the open-loop framework, the closed-loop engine controller can be designed using a decentralized structure without significantly affecting the system performance. In the next step, based on a nonlinear physics-based model of the engine, a set of single-input single-output (SISO) adaptive second order discrete sliding mode controllers (DSMC) are designed to drive the states of the engine model to their pre-defined desired trajectories and minimize the tailpipe HC emission, under modelling and implementation (data sampling and quantization) uncertainties. The real-time test results on an actual engine control unit (ECU) show that the proposed SISO adaptive second order DSMC provides accurate and fast tracking performance for the highly nonlinear and internally coupled engine dynamics, and can meet the HC emission limit by controlling the engine-out emissions and exhaust catalytic converter efficiency., Comment: 22 pages, 10 figures, 2017 International Conference on Advanced Vehicle Powertrains

The performance of a conventional model-based controller significantly depends on the accuracy of the modeled dynamics. The model of a plant's dynamics is subjected to errors in estimating the numerical values of the physical parameters, and variations over operating environment conditions and time. These errors and variations in the parameters of a model are the major sources of uncertainty within the controller structure. Digital implementation of controller software on an actual electronic control unit (ECU) introduces another layer of uncertainty at the controller inputs/outputs. The implementation uncertainties are mostly due to data sampling and quantization via the analog-to-digital conversion (ADC) unit. The failure to address the model and ADC uncertainties during the early stages of a controller design cycle results in a costly and time consuming verification and validation (V&V) process. In this paper, new formulations of the first and second order discrete sliding mode controllers (DSMC) are presented for a general class of uncertain linear systems. The knowledge of the ADC imprecisions is incorporated into the proposed DSMCs via an online ADC uncertainty prediction mechanism to improve the controller robustness characteristics. Moreover, the DSMCs are equipped with adaptation laws to remove two different types of modeling uncertainties (multiplicative and additive) from the parameters of the linear system model. The proposed adaptive DSMCs are evaluated on a DC motor speed control problem in real-time using a processor-in-the-loop (PIL) setup with an actual ECU. The results show that the proposed SISO and MIMO second order DSMCs improve the conventional SISO first order DSMC tracking performance by 69% and 84%, respectively. Moreover, the proposed adaptation mechanism is able to remove the uncertainties in the model by up to 90%., Comment: 10 pages, 11 figures, ASME 2017 Dynamic Systems and Control Conference

Analog-to-digital conversion (ADC) and uncertainties in modeling the plant dynamics are the main sources of imprecisions in the design cycle of model-based controllers. These implementation and model uncertainties should be addressed in the early stages of the controller design, otherwise they could lead to failure in the controller performance and consequently increase the time and cost required for completing the controller verification and validation (V&V) with more iterative loops. In this paper, a new control approach is developed based on a nonlinear discrete sliding mode controller (DSMC) formulation to mitigate the ADC imprecisions and model uncertainties. To this end, a DSMC design is developed against implementation imprecisions by incorporating the knowledge of ADC uncertainties on control inputs via an online uncertainty prediction and propagation mechanism. Next, a generic online adaptive law will be derived to compensate for the impact of an unknown parameter in the controller equations that is assumed to represent the model uncertainty. The final proposed controller is an integrated adaptive DSMC with robustness to implementation and model uncertainties that includes (i) an online ADC uncertainty mechanism, and (ii) an online adaptation law. The proposed adaptive control approach is evaluated on a nonlinear automotive engine control problem in real-time using a processor-in-the-loop (PIL) setup with an actual electronic control unit (ECU). The results reveal that the proposed adaptive control technique removes the uncertainty in the model fast, and significantly improves the robustness of the controllers to ADC imprecisions. This provides up to 60% improvement in the performance of the controller under implementation and model uncertainties compared to a baseline DSMC, in which there are no incorporated ADC imprecisions., Comment: 10 pages, 10 figures, ASME 2016 Dynamic Systems and Control Conference (DSCC)

Sliding mode control (SMC) is a robust and computationally efficient solution for tracking control problems of highly nonlinear systems with a great deal of uncertainty. High frequency oscillations due to chattering phenomena and sensitivity to data sampling imprecisions limit the digital implementation of conventional first order continuous-time SMC. Higher order discrete SMC is an effective solution to reduce the chattering during the controller software implementation, and also overcome imprecisions due to data sampling. In this paper, a new adaptive second order discrete sliding mode control (DSMC) formulation is presented to mitigate data sampling imprecisions and uncertainties within the modeled plant's dynamics. The adaptation mechanism is derived based on a Lyapunov stability argument which guarantees asymptotic stability of the closed-loop system. The proposed controller is designed and tested on a highly nonlinear combustion engine tracking control problem. The simulation test results show that the second order DSMC can improve the tracking performance up to 80% compared to a first order DSMC under sampling and model uncertainties., Comment: 6 pages, 6 figures, 2017 American Control Conference