Search Results

To prevent the severe consequences of unsafe driving behaviors, it is crucial to monitor and analyze the state of the driver. Developing an effective driver state monitoring (DSM) systems is particularly challenging due to limited computation capabilities of embedded systems in automobiles and the need for finishing processing in real-time. However, most of the existing research work was conducted in a lab environment with expensive equipment while lacking in-car benchmarking and validation. In this paper, a DSM system that estimates driver's alertness and drowsiness level as well as performs emotion detection built with a cost-effective embedded system is presented. The proposed system consists of a mono camera that captures driver's facial image in real-time and a machine learning based detection algorithm that detects facial landmark points and use that information to infer driver's state.

As the piston pin works under significant mechanical load, it is susceptible to wear, seizure, and structural failure, especially in heavy duty internal combustion engines. It has been found that the friction loss associated with the pin is comparable to that of the piston, and can be reduced when the interface geometry is properly modified. However, the mechanism that leads to such friction reduction, as well as the approaches towards further improvement, remain unknown. This work develops a piston pin lubrication model capable of simulating the interaction between the pin, the piston, and the connecting rod. The model integrates dynamics, solid contact, oil transport, and lubrication theory, and applies an efficient numerical scheme with second order accuracy to solve the highly stiff equations. As a first approach, the current model assumes every component to be rigid.

A 1-dimensional analytic solution has been developed to evaluate the pressure drop and filtration performance of ceramic wall-flow partial diesel particulate filters (PFs). An axially resolved mathematical model for the static pressure and velocity profiles prevailing inside wall-flow filters, with such unique plugging configurations, is being proposed for the first time. So far, the PF models that have been developed are either iterative/numerical in nature [1], or based on commercial CFD packages [7]. In comparison, an analytic solution approach is a transparent and computationally inexpensive tool that is capable of accurately predicting trends as well as, offering explanations to fundamental performance behavior. The simple mathematical expressions that have been obtained facilitate rational decision-making when designing partial filters, and could also reduce the complexity of OBD logic necessary to control onboard filter performance.

Intense competition in the automotive industry requires continuous reduction in innovation cycle time, even as corporations are downsizing and system complexity is increasing. Subsequently, the application of recently introduced Rapid Algorithm Development (RAD) tools has facilitated significant advances in the development of embedded control systems. The RAD steps include system modeling, control algorithm design, simulation analysis, automated calibration design, and vehicle implementation through automatic code generation. The application of RAD tools and the associated benefits are described, specifically in the context of Engine Management Systems (EMS). Such benefits include significant reductions in development cycle time, open architecture, automated calibration, and information reuse.

An extravehicular activity (EVA) path-planning and navigation tool, called the Mission Planner, has been developed to assist with pre-mission planning, scenario simulation, real-time navigation, and contingency replanning during astronaut EVAs, The Mission Planner calculates the most efficient path between user-specified waypoints. Efficiency is based on an exploration cost algorithm, which is a function of the estimated astronaut metabolic rate. Selection of waypoints and visualization of the generated path are realized within a 3D mapping interface through terrain elevation models. The Mission Planner is also capable of computing the most efficient path back home from any point along the path.

This paper presents the business purpose, software architecture, technology integration, and applications of the Cummins Vehicle Mission Simulation (VMS) software. VMS is the value-based analysis tool used by the marketing, sales, and product engineering functions to simulate vehicle missions quickly and to gauge, communicate, and improve the value proposition of Cummins engines to customers. VMS leverages the best of software architecture practices and proven technologies available today. It consists of a close integration of MATLAB and Simulink with Java, XML, and JDBC technologies. This Windows compatible application software uses stand-alone mathematical models compiled using Real Time Workshop. A built-in MySQL database contains product data for engines, driveline components, vehicles, and topographic routes. This paper outlines the database governance model that facilitates effective management, control, and distribution of engine and vehicle data across the enterprise.

This paper presents a brake pressure estimation algorithm for Delphi Traction Control Systems (TCS). A control oriented lumped parameter model of a brake control system is developed using Matlab/Simulink. The model is derived based on a typical brake system and is generic to other types of brake control hardware systems. For application purposes, the model is simplified to capture the dominant dynamic brake pressure response. Vehicle experimental data collected under various scenarios are used to validate the algorithm. Simulation results show that the algorithm gives accurate pressure estimation. In addition, the calibration procedure is greatly simplified

This paper presents a theoretical and experimental study of a cylinder deactivation valvetrain system for the integration into an Engine Management System (EMS). A control-oriented lumped parameter model of the deactivation valvetrain system is developed and implemented using Matlab/Simulink, and validated by experimental data. Through simulation and experimental data analysis, the effect of operating conditions on the dynamic response is captured and characterized, over a wide range of operating conditions. The algorithm provides a basis for the calibration of the deactivation hardware. The generic characterization of the dynamic response can simplify the calibration parameters for the implementation in engine management systems.

This paper describes an environment for the development of production Engine Management Systems (EMS). This includes a formal framework and modeling methodology. The environment is based on using Simulink/Stateflow for developing a control system executable specification and a plant model. This allows for simulations of the system to be performed at the engineer's desk, which is identical performance with production software. We provide the details for incorporating production legacy code into the Simulink/Stateflow control system. The system includes a multi-rate, and event driven operating system. This system is developed to facilitate new algorithm development and automated software testing. Based on Simulink/Stateflow this specification will be suitable for use with commercial automatic code generation tools.

This paper describes the verification and comparison of position control algorithms for a continuously-variable cam phaser. Robust Engineering techniques are used. Two non-linear PID control algorithms are designed to control cam phaser position. The first algorithm is a more complex control strategy while the second is a thrifted approach that seeks to reduce throughput requirements. An L18 orthogonal array is established with noise factors that affect the quality of cam phaser control. Using the orthogonal array, the number of experiment test points required to characterize the control algorithm response is reduced from 8,748 to thirty-six. The test points of the orthogonal array are investigated experimentally on a motored engine outfitted with cam phaser hardware. The desired and actual cam position data are compared and analyzed for all points in the orthogonal array.

This paper describes the development and implementation of an Engine Drag Control algorithm to improve vehicle stability performance. Engine drag can occur on low and high coefficient surfaces when the driver suddenly releases the throttle. If the engine drag force becomes larger than the frictional force between the tire and the road, the tires will break loose from the surface and slip. This could induce vehicle instability especially with rear drive vehicles on low-coefficient surfaces. The EDC algorithm has been developed to provide accurate control of the wheels. EDC will help reduce the yaw rate of the vehicle and thus achieve greater vehicle stability. The paper also presents methods used to test the robustness of such a system. The purpose of the testing was to ensure that there would be no false activations of EDC under normal driving conditions and also to ensure that, when the system is active, it is mostly transparent to the driver.

Today automotive system suppliers develop more-or-less independent systems, such as brake, power steering and suspension systems. In the future, car manufacturers like Volvo will build up vehicle control systems combining their own algorithms with algorithms provided by automotive system suppliers. Standardization of interfaces to actuators, sensors and functions is an important enabler for this vision and will have major consequences for functionality, prices and lead times, and thus affects both vehicle manufacturers and automotive suppliers. The investigation of the level of appropriate interfaces, as part of the European BRAKE project, is described here. Potential problems and consequences are discussed from both a technical and a business perspective. This paper provides a background on BRAKE and on the functional decomposition upon which the interface definitions are based. Finally, the interface definitions for brake system functionality are given.

Delphi is developing a new combustion technology called Gasoline Direct-injection Compression Ignition (GDCI), which has shown promise for substantially improving fuel economy. This new technology is able to reuse some of the controls common to traditional spark ignition (SI) engines; however, it also requires several new sensors and actuators, some of which are not common to traditional SI engines. Since this is new technology development, the required hardware set has continued to evolve over the course of the project. In order to support this development work, a highly capable and flexible electronic control system is necessary. Integrating all of the necessary functions into a single controller, or two, would require significant up-front controller hardware development, and would limit the adaptability of the electronic controls to the evolving requirements for GDCI.

A Multi-Objective Optimization (MOO) problem concerning the thermal control problem of Multifunctional Structures (MFSs) is here addressed. In particular the use of Multi-Objective algorithms from an optimization tool and Self-Organizing Maps (SOM) is proposed for the identification of the optimal topological distribution of the heating components for a multifunctional test panel, the Advanced Bread Board (ABB). MFSs are components that conduct many functions within a single piece of hardware, shading the clearly defined boundaries that identify traditional subsystems. Generally speaking, MFSs have already proved to be a disrupting technology, especially in aeronautics and space application fields. The case study exploited in this paper refers to a demonstrator breadboard called ABB. ABB belongs to a particular subset of an extensive family of MFS, that is, of thermo-structural panels with distributed electronics and a health monitoring network.

Human explorers of planetary surfaces would benefit greatly from a spacesuit design that facilitates locomotion. To aid in the development of such an extravehicular activity suit, a design effort incorporating the concept of mechanical counter pressure (MCP) was undertaken. Three-dimensional laser scanning of the human body was used to identify the main effects of knee flexion angle on the size and shape of the leg. This laser scanning quantified the changes in shape that must be supported by an MCP garment and the tension that must be developed to produce even MCP. Evaluation of a hybrid-MCP concept using inextensible materials demonstrated strong agreement between experimental data and a mathematical model with rigid cylinder geometry. Testing of a form-fitting garment on the right lower leg of a subject demonstrated successful pressure production. Further research is required to evaluate how evenly pressure can be distributed using the hybrid-MCP concept.

An algorithm for estimation of vehicle yaw rate and side slip angle using steering wheel angle, wheel speed, and lateral acceleration sensors is proposed. It is intended for application in vehicle stability enhancement systems, which use controlled brakes or steering. The algorithm first generates two initial estimates of yaw rate from wheel speeds and from lateral acceleration. A new estimate is subsequently calculated as a weighted average of the two initial ones, with the weights proportional to confidence levels in each estimate. This preliminary estimate is fed into a closed loop nonlinear observer, which generates the final estimate of yaw rate along with estimates of lateral velocity and side slip angle. Parameters of the observer depend on the estimated surface coefficient of adhesion, thus providing adaptation to changes in road surface coefficient of adhesion.

A finite element model of a folded airbag with the module cover and steering wheel system was developed to estimate the injury numbers of a 5th percentile female dummy in an out-of-position (OOP) situation. The airbag model was correlated with static airbag deployments and standard force plate tests. The 5th percentile finite element dummy model developed by First Technology Safety Systems (FTSS) was used in the simulation. The following two OOP tests were simulated with the airbag model including a validated steering wheel finite element model: 1. Chest on air bag module for maximum chest interaction from pressure loading (MS6-D) and 2. Neck on air bag module for maximum neck interaction from membrane loading (MS8-D). These two simulations were then compared to the test results. Satisfactory correlation was found in both the cases.

Automotive occupant safety continues to evolve. At present this area has gathered a strong consumer interest which the vehicle manufacturers are tapping into with the introduction of many new safety technologies. Initially, individual passive devices and features such as seatbelts, knee- bolsters, structural crush zones, airbags etc., were developed for to help save lives and minimize injuries in accidents. Over the years, preventive measures such as improving visibility, headlights, windshield wipers, tire traction etc., were deployed to help reduce the probability of getting into an accident. With tremendous new research and improvements in electronics, we are at the stage of helping to actively avoid accidents in certain situations as well as providing increased protection to vehicle occupants and pedestrians.

A Barometric Pressure Estimator (BPE) algorithm was implemented in a production speed-density Engine Management System (EMS). The BPE is a model-based, easily calibrated algorithm for estimating barometric pressure using a standard set of production sensors, thereby avoiding the need for a barometric pressure sensor. An accurate barometric pressure value is necessary for a variety of engine control functions. By starting with the physics describing the flow through the induction system, an algorithm was developed which is simple to understand and implement. When used in conjunction with the Pneumatic and Thermal State Estimator (PSE and TSE) algorithms [2], the BPE requires only a single additional calibration table, generated with an automated processing routine, directly from measured engine data collected at an arbitrary elevation, in-vehicle or on a dynamometer. The algorithm has been implemented on several different engines.

In this paper we discuss in detail an algorithm that addresses cylinder-to-cylinder imbalance issues. Maintaining even equivalence-ratio (ϕ) control across all the cylinders of an engine is confounded by imbalances which include fuel-injector flow variations, fresh-air intake maldistribution and uneven distribution of Exhaust Gas Re-circulation (EGR). Moreover, in markets that are growing increasingly cost conscious, with ever tightening emissions regulations, correcting for such mismatches must not only be done, but done at little or no additional cost. To address this challenge, we developed an Individual Cylinder Fuel Control (ICFC) algorithm that estimates each cylinder's individual ϕ and then compensates to correct for any imbalance using only existing production hardware. Prior work in this area exists1,2, yet all disclosed production-intent work was performed using wide-range oxygen sensors, representing cost increases.