Fazal Urrahman Syed

Derivation and Experimental Validation of a Power-Split Hybrid Electric Vehicle Model

Hybrid electric vehicles (HEVs) have attracted a lot of attention due to environmental and efficiency reasons. Typically, an HEV combines two power trains, a conventional power source such as a gasoline engine, a diesel engine, or a fuel cell stack, and an electric drive system (involving a motor and a generator) to produce driving power with a potential of higher fuel economy than conventional vehicles. Furthermore, such vehicles do not require external charging and thus work within the existing fueling infrastructures. The power-split power train configuration of an HEV has the individual advantages of the series and parallel types of HEV power train configurations. A sophisticated control system, however, is required to manage the power-split HEV power trains. Designing such a control system requires a reasonably accurate HEV system plant model. Much research has been done for developing dynamic plant models for the series and parallel types, but a complete and validated dynamic model for the power-split HEV power train is still in its infancy. This paper presents a power-split power train HEV dynamic model capable of realistically replicating all the major steady-state and transient phenomena appearing under different driving conditions. A mathematical derivation and modeling representation of this plant model and its components is shown first. Next, the analysis, verification, and validation through computer simulation and comparison with the data actually measured in the test vehicle at the Ford Motor Companys test track is performed. The excellent agreements between the model and the experimental results demonstrate the fidelity and validity of the derived plant model. Since this plant model was built by integrating the subsystem models using a system-oriented approach with a hierarchical methodology, it is easy to change subsystem functionalities. The developed plant model is useful for analyzing and understanding the dominant dynamics of the power train system, the interaction between subsystems and components, and system transients due to the change of operational state and the influence of disturbances. This plant model can also be employed for the development of vehicle system controllers, evaluation of energy management strategies, issue resolution, and verification of coded algorithms, among many other purposes

Fuzzy Gain-Scheduling Proportional–Integral Control for Improving Engine Power and Speed Behavior in a Hybrid Electric Vehicle

With the increased emphasis on improving fuel economy and reducing emissions, hybrid electric vehicles (HEVs) have emerged as very strong candidates to achieve these goals. The power-split hybrid system, which is a complex hybrid powertrain, exhibits great potential to improve fuel economy by determining the most efficient regions for engine operation and thereby high-voltage (HV) battery operation to achieve overall vehicle efficiency optimization. To control and maintain the actual HV battery power, a sophisticated control system is essential, which controls engine power and thereby engine speed to achieve the desired HV battery maintenance power. Conventional approaches use proportional-integral (PI) control systems to control the actual HV battery power in power-split HEV, which can sometimes result in either overshoots of engine speed and power or degraded response and settling times due to the nonlinearity of the power-split hybrid system. We have developed a novel approach to intelligently controlling engine power and speed behavior in a power-split HEV using the fuzzy control paradigm for better performances. To the best of our knowledge, this is the first reported use of the fuzzy control method to control engine power and speed of a power-split HEV in the applied automotive field. Our approach uses fuzzy gain scheduling to determine appropriate gains for the PI controller based on the systems operating conditions. The improvements include elimination of the overshoots as well as approximate 50% faster response and settling times in comparison with the conventional linear PI control approach. The improved performances are demonstrated through simulations and field experiments using a ford escape hybrid vehicle.

Active Damping Wheel-Torque Control System to Reduce Driveline Oscillations in a Power-Split Hybrid Electric Vehicle

Power-split hybrid electric vehicles (HEVs) provide a great opportunity to improve fuel economy and emissions. This power-split hybrid system has inherent low damping in driveline since it uses planetary gear sets to directly connect the engine, the generator, and the motor to the driveline for improved vehicle efficiency, thus lacking a clutch or a torque converter that provides the conventional vehicles with driveline damping. When they are subjected to acceleration or disturbances, the low damping in the driveline may cause torsional vibrations. Since the power-split control system is closed loop in nature, these torsional vibrations can result in sustained driveline oscillations. These oscillations can be very objectionable to the driver as they affect the vehicles drivability. In this paper, we present the design of an active damping wheel-torque control system to suppress such oscillations to improve the drivability of a power-split HEV. To the best of our knowledge, this is the first reported use of an active damping wheel-torque control system to suppress the driveline oscillations in a power-split HEV. Simulations in a power-split HEV environment and experimental tests in the field using a Ford Escape Hybrid demonstrate the effectiveness of the proposed system in suppressing the oscillations. The driveline disturbances are suppressed to below the perceptible level of wheel torque (<100 Nmiddotm). Additional simulations are performed to validate the system to other key factors that can affect its performance. Even with increased motor/generator disturbances by a factor of 2 and change in driveline stiffness of plusmn50%, the proposed control system can still effectively suppress driveline oscillations and thereby improve drivability.

Fuzzy control to improve high-voltage battery power and engine speed control in a hybrid electric vehicle

With the recent emphasis on developing more environmentally friendly and fuel-efficient vehicles, Ford Motor Company developed a full hybrid electric vehicle (HEV) with a power-split hybrid powertrain consisting of an integrated motor and a generator. The power-split hybrid consists of two powertrains; an engine and an electric drive system. This powertrain provides a great potential to improve fuel economy in part due to its ability to operate engine at efficient regions independent of the vehicle speed. The engine speed determination in such a system depends on the desired high voltage (HV) battery power and the driver demand (driver torque/power request). Clearly, in order to control HV battery power to a desired power, a sophisticated controls system is essential which controls engine power to achieve the desired HV battery power. The desired engine power in turn determines the desired engine speed. It is essential that engine speed operation is smooth and stable with an acceptable response. Use of a classical proportional-integral (PI) based control system to control HV battery power is limited due to the nonlinear behavior of the powertrain, and results either in an undesired engine speed stability behavior under certain driving conditions or degraded response time. This paper presents a new nonlinear controls scheme based on a fuzzy controller to resolve the undesired engine speed behavior while achieving desired engine speed response and improved high-voltage battery power controls. Simulations are conducted with this controller and results show that the proposed fuzzy controller improves HV battery power controls and thereby the engine speed behavior and response time (e.g., no overshoots, improved settling time, and uncompromised rise time).

Adaptive real-time advisory system for fuel economy improvement in a hybrid electric vehicle

In this paper, we present a fuzzy logic based adaptive algorithm with a learning mechanism that estimates drivers long term and short term preferences. The algorithm represents a significant advancement to the capability of our previous non-adaptive real-time fuel economy advisory system that was implemented in a Ford Escape Hybrid [8][9]. This real-time advisory system proposed in [8][9]achieved improved fuel economy by providing visual and haptic feedbacks to the driver to change his or her driving style or behavior for a given vehicle condition. It was tuned to maximize fuel economy without significantly impacting the performance of the vehicle. Some drivers may perceive its feedback to be intrusive on one extreme while some other drivers may feel it ineffective on another extreme, depending on the drivers driving styles. The new adaptive algorithm learns drivers intentions by monitoring their driving styles and behaviors, and addresses the issues of intrusiveness of the advisory feedback. This proposed adaptive algorithm balances the competing requirements for improved fuel economy and drivability by maintaining vehicle performance that is acceptable to the current drivers driving style and behavior while providing mechanism to improve fuel economy. This system was developed and validated on the Ford Escape Hybrid vehicle. Experimental results show that the proposed adaptive algorithm is capable of improving drivers behavior and style without being perceived as ineffective or intrusive and achieves fuel economy improvements.

Applied intelligent systems: blending fuzzy logic with conventional control

The aim of this paper is to show that design of applied intelligent control systems requires different types of blending between fuzzy logic and conventional control systems. Two alternative automotive applications – a manufacturing process control problem and an advisory system for fuel efficient driving – that benefit from both fuzzy and control theories are reviewed and different levels of prioritisations of both approaches are discussed based on the specificity of the applications.

Rule-Based Fuzzy Gain-Scheduling PI Controller to Improve Engine Speed and Power Behavior in a Power-split Hybrid Electric Vehicle

Environmental awareness has resulted in greater emphasis on developing more environmentally friendly and fuel efficient vehicles. Hybrid electric vehicles (HEVs) have been considered a viable option towards achieving these goals. Ford Motor Company developed a full hybrid electric vehicle with an e-CVT (electronically controlled continuously variable transmission) or power-split hybrid powertrain with an integrated motor and generator. The power-split hybrid system uses planetary gear sets to connect an engine, a generator, and a motor. This HEV powertrain exhibits great potential to improve fuel economy by enabling the engine to operate at its most efficient region independent of the vehicle speed. To achieve fuel economy improvements of the power-split hybrid system, high-voltage (HV) battery power management is critical. To control actual HV battery power in such vehicles, a sophisticated control system is essential which controls engine power and thereby engine speed to achieve the desired HV battery maintenance power. Conventional approaches use proportional-integral (PI) control systems to control the actual HV battery power in power-split hybrid system, which can sometimes result in either overshoots of engine speed and power or degraded response and settling times due to the nonlinearity of the power-split hybrid system. Such an overshoot is often objectionable to customers, which see engine speed overshoots as disconnect between the drivers request and the engine response. This issue comes from the fact that a complete high fidelity mathematical model for the power-split HEV system along with the environmental effects cannot be accurately modeled inside the controller. Therefore, a controller adaptable to nonlinear behavior and not requiring detailed knowledge of mathematical model of the plant is required to address such issues. Fuzzy control approaches can provide a way to cope with the limitations of the conventional controllers. We have developed a fuzzy control approach with minimal rules to intelligently control engine power and speed behavior in a powersplit HEV. This approach uses selective minimal rule-based fuzzy gain-scheduling to determine appropriate gains for the PI controller based on the systems operating conditions. The improvements result in the reduction of the overshoots without compromising systems response and settling times in comparison with the conventional linear PI controller. This paper describes the power-split hybrid vehicles powertrain system and key subsystems. It also describes minimal rule-based fuzzy gain-scheduling PI controller and the formulation of minimal fuzzy rules required to achieve the desired behavior. This minimal rule-based fuzzy controller was implemented in a Ford Escape hybrid vehicle and was evaluated in the vehicle test environment for a comparative analysis of the results to show its effectiveness. The results clearly demonstrate that the designed minimal rule based fuzzy gains scheduling controller is capable of significantly improving the engine speed and power behavior in a power-split HEV without compromising the systems response and settling times

A two-stage-training support vector machine approach to predicting unintentional vehicle lane departure

ABSTRACT Advanced driver assistance systems, such as unintentional lane departure warning systems, have recently drawn much attention and efforts. In this study, we explored utilizing the nonlinear binary support vector machine (SVM) technique to predict unintentional lane departure, which is innovative, as the SVM methodology has not previously been attempted for this purpose in the literature. Furthermore, we developed a two-stage training scheme to improve SVMs prediction performance in terms of minimization of the number of false positive prediction errors. Experiment data generated by VIRTTEX, a hydraulically powered, 6-degrees-of-freedom moving base driving simulator at Ford Motor Company, were used. All the vehicle variables were sampled at 50 Hz and there were 16 drowsy drivers (about 3 hours of driving per subject) and six control drivers (approximately 20 minutes f driving each). In total, 3,508 unintentional lane departures occurred for the drowsy drivers and 23 for the control drivers. Our study involving these 22 drivers with a total of more than 7.5 million prediction decisions demonstrates that (a) excellent SVM prediction performance, measured by numbers of false positives (i.e., falsely predicted lane departures) and false negatives (i.e., lane departures failed to be predicted), was achieved when the prediction horizon was 0.6 seconds or less, (b) lateral position and lateral velocity worked the best as SVM input variables among the nine variable sets that we explored, and (c) the radial basis function performed the best as the SVM kernel function.

A support vector machine approach to unintentional vehicle lane departure prediction

Advanced driver assistance systems, such as unintentional lane departure warning systems, have recently drawn much attention and R & D efforts. Such a system may assist the driver by monitoring the driver or vehicle behaviors to predict/detect driving situations (e.g, lane departure) and alert the driver to take corrective action. In this paper, we show how the support vector machine (SVM) methodology can potentially provide enhanced unintentional lane departure prediction, which is a new method relative to literature. Our binary SVM employed the Radial Basis Function kernel to classify time series of select vehicle variables. The SVM was trained and tested using the driver experiment data generated by VIRTTEX, a hydraulically powered 6-degrees-of-freedom moving base driving simulator at Ford Motor Company. The data that we used represented 16 drowsy subjects (three-hour driving time per subject) and six control subjects (20 minutes driving per subject), all of which drove a simulated 2000 Volvo S80. The vehicle variables were all sampled at 50 Hz. There were a total of 3,508 unintentional lane departure occurrences for the drowsy drivers and only 23 for four of the six control drivers (two had none). The SVM was trained by over 60,000 time series examples (the actual number depended on the prediction horizon) created from 50% of the lane departures. The training data were removed from the testing data. During the testing, the SVM made a lane departure prediction at every sampling time for every one of the 22 drivers (over 6.8 million predictions in total). The overall sensitivity and specificity of the SVM with a 0.2-second prediction horizon for the 22 drivers were 99.77465% and 99.99997%, respectively. The SVM predicted, on average 0.200181 seconds in advance, lane departure correctly for all the control drivers, but missed 4 of the 1,758 and gave false positives for another 2 for the drowsy drivers. For the prediction horizon of 0.4s, there was 1 false positive case for the control subjects, and the false negative and false positive cases rose substantially to 10 and 137 for the drowsy drivers, respectively.

A neural network for predicting unintentional lane departures

Unintended lane departure accidents are due to drivers inattention, incapacitation, and drowsiness. Lane departure warning systems have been developed to enhance traffic safety by predicting/detecting driving situation and alerting drivers to avoid or mitigate traffic accidents. This paper explores effectiveness of a three-layer perceptron neural network in predicting an unintentional lane departure, which to the best of our knowledge has not been reported in the literature. This study used driver experiment data generated by VIRTTEX, a hydraulically powered 6-degrees-of-freedom moving base driving simulator at Ford Motor Company. The experimental data represented 16 drowsy drivers who drove a simulated 2000 Volvo S80 (three hours per driver), which consisted of a total of 3,508 lane departure occurrences. Two-third of the lane departures were randomly selected to generate training examples for the network (82,040 examples for a 0.2-second prediction horizon and 171,112 for a 0.5-second horizon). The number of hidden neurons as well as the input vehicle variables were optimized experimentally through the training process. The optimized network was then used to predict lane departure by processing the entire driving time series of the 16 drivers one by one after all the training data was removed from the time series. The network made a prediction at each sampling moment of the time series and there were over 6.3 million predictions. The overall recall and precision of the optimized network for the 0.2-second horizon were 99.74% and 99.66%, respectively, which degraded to 99.23% and 85.49%, respectively, when the horizon increased to 0.5 s.