Search Results

The aim of this work is to apply an innovative adaptive ℒ1 techniques to control flutter phenomena affecting highly flexible wings and to evaluate the efficiency of this control algorithm and architecture by performing the following tasks: i) adaptation and analysis of an existing simplified nonlinear plunging/pitching 2D aeroelastic model accounting for structural nonlinearities and a quasi-steady aerodynamics capable of describing flutter and post-flutter limit cycle oscillations, ii) implement the ℒ1 adaptive control on the developed aeroelastic system to perform initial control testing and evaluate the sensitivity to system parameters, and iii) perform model validation and calibration by comparing the performance of the proposed control strategy with an adaptive back-stepping algorithm. The effectiveness and robustness of the ℒ1 adaptive control in flutter and post-flutter suppression is demonstrated.

Many newer commercial vehicles have an event data recorder (EDR) that can record pre-event and post-event speeds. The EDR is incorporated into the engines electronic control module (ECM). In this study, the accuracy of the ECM-reported speed was tested during acceleration, gear shifting and braking at speeds between 16 and 88 km/h (10 to 55mph). The ECM-reported speed was compared to the speed measured by a calibrated optical 5th wheel. The results showed that the accuracy of the ECM-reported speed matched closely during acceleration, cycled to periods of under-reporting the speed during hard braking due to the ABS brake function, briefly under-reporting the speed after letting off the throttle for braking or gear shift and briefly over-reporting the speed near the end of a gear shift phase. This study also looked at calibration factors of the ECM and their effect on the ECM-reported speed.

The onboard “SODART” telescope silicon detector cooling system of the “Spectrum-X-Gamma” observatory, which is designed for the space objects X-ray radiation study, is described. The scale-down model of the passive cooling system description and thermal vacuum test results of this model are given. In the real cooling system the minimal detector temperature at 300 mW heat release is expected about 107 K.

Mathematical modeling and control of artificial ecosystems, such as MELISSA, require first the study of physical and biological characteristics in optimal and limiting conditions. Following the previous determination of the stoichiometric equations (Spirulina compartment) and regarding the two phototrophic compartments of MELISSA (Rhodospirillaceae and Spirulina), we have first to focus our control study on the growth kinetics for the light source. In this paper, we recall the theoretical equations of microbial growth kinetics and emphasise the problem of the light transfer in a photobioreactor. We present their adaptations to our pilot plant taking into account technological and biological specifics (lamp spectrum, working illuminated volume, growth rate,…). We then develop the principles and structure of the control system and describe tests of both the hardware and software for several steady state configurations.

This paper describes the application of statistical techniques known as Design of Experiments (D.O.E.) to efficiently use the results of numerical analysis data in order to improve the configuration of automotive systems. The general framework of these techniques is presented in a format aiming at the design engineer as their end user. Besides, a case study is presented with the purpose of illustrating their practical use. The first step of the case study is to build predictive models for the behaviour of the automotive system being developed by means of the Response Surface Method (RSM), using the proper D.O.E. options. Once these predictive models are available, automatic numerical optimization algorithms are used to improve the responses of interest for given operating conditions. Finally, the automotive systems are robust designed taking into account that the operating conditions vary randomly.

Driver state monitoring is a crucial technology for enhancing road safety and preventing human error-caused accidents in the era of autonomous vehicles. This paper presents CogniSafe, a comprehensive driver monitoring system that uses deep learning and computer vision methods to detect various types of driver distractions and fatigue. CogniSafe consists of four modules: Driver anomaly detection and classification: A novel two-phase network that proposes and recognizes driver anomalies, such as texting, drinking, and adjusting radios, using multimodal and multiview input. Gaze estimation: A video-based neural network that jointly learns head pose and gaze dynamics, achieving robust and efficient gaze estimation across different head poses. Eye state analysis: A multi-tasking CNN that encodes features from both eye and mouth regions, predicting the percentage of eye closure (PERCLOS) and the frequency of mouth opening (FOM).

This paper is an extension of our previous work on the CHASE (Classification by Holistic Analysis of Scene Environment) algorithm, that automatically classifies the driving complexity of a road scene image during day-time conditions and assigns it an ‘Ease of Driving’ (EoD) score. At night, apart from traffic variations and road type conditions, illumination changes are a major predominant factor that affect the road visibility and the driving easiness. In order to resolve the problem of analyzing the driving complexity of roads at night, a brightness detection module is incorporated in our end-to-end nighttime EoD system, which computes the ‘brightness factor’ (bright or dark) for that given night-time road scene. The brightness factor along with a multi-level machine learning classifier is then used to classify the EoD score for a night-time road scene. Our end-to-end ‘Night-time EoD system’ is a real-time onboard system implemented and tested on road scene data collected in Japan.

The U.S. National Highway Transportation and Safety Agency's (NHTSA) early estimates of Motor Traffic Fatalities in 2009 in the United States [1] show continuing progress on improving traffic safety on the U.S. roadways. The number of total fatalities and the fatality rate per 100 Million Vehicle Miles (MVM), both show continuing declines. In the 10 year period from 1999 through 2009, the total fatalities have dropped from 41,611 to 33,963 and the fatality rate has dropped from 1.5 fatalities per 100MVM to 1.16 fatalities per 100MVM, a compound annual drop of 2.01% and 2.54% respectively. The large number of traffic fatalities, and the slowing down of the fatality rate decline, compared to the decade before, continues to remain a cause of concern for regulators.

The objective of the development of the aerodynamic drag predictive tool CDaero was for use as a module for the Automobile Design Support System (AutoDSS). CDaero is an empirically based drag coefficient predictive tool based initially on the MIRA (Motor Industry Research Association) algorithm. The development philosophy was to be able to predict the aerodynamic drag coefficient of an automobile with knowledge of the features of the surface geometry control curves. These are the curves that control the 3-dimensional geometry as seen in the profile, plan and front and rear views. CDaero has been developed in a computing environment using the equation solver TKSolver™. Fifty-one input feature values are first determined from the automobile geometry and then entered into the program. CDaero models the drag coefficient with thirteen different components covering the basic body, as well as additional components such as the wheels, mud flaps, etc.

This is an overview of air bag injuries, a review of the literature and descriptions of air bag related injuries to the various body areas. Some unusual injuries are also included. The cases presented are from the author's files and one from the Special Studies Division of NHTSA.

The Swedish Biomimetics 3000's μMist® platform technology has been used to develop a radically new injection system. This prototype system, developed and characterized with support from Lotus, as part of Swedish Biomimetics 3000®'s V₂IO innovation accelerating model, delivers improved combustion efficiency through achieving exceptionally small droplets, at fuel rail pressures far less than conventional GDI systems and as low as PFI systems. The system gives the opportunity to prepare and deliver all of the fuel load for the engine while the intake valves are open and after the exhaust valves have closed, thereby offering the potential to use advanced charge scavenging techniques in PFI engines which have hitherto been restricted to direct-injection engines, and at a lower system cost than a GDI injection system.

Propulsion system control algorithms covering the functional needs of xEV propulsion (‘x’ donates P0-P4 configurations) systems are presented in this paper. The scope and foundation are based on generic well-established HEV controller architectures. However, unlike conventional HEV (series, parallel and power split) powertrains, the next generation of integrated electric propulsion configurations will utilize a single micro controller that supports multiple control functions ranging from the electric machines, inverters, actuators, clutch solenoids, coolant pumps, etc. This presents a unique challenge to architect control algorithms within the AUTOSAR framework while satisfying the complex timing requirements of motor/generator-inverter (MGi) control and increased interface definitions between software components to realize functional integration between the higher level propulsion system and its sub-systems.

eROSITA (extended ROentgen Survey with an Imaging Telescope Array) is a powerful X-ray telescope under development by the Max-Planck-Institut für extraterrestrische Physik (MPE) in Garching, Germany. eROSITA is the core instrument on the Russian SRG1 mission which is planned for launch in 2011. It comprises seven nested Wolter-I grazing incidence telescopes, each equipped with its own CCD camera. The mirror modules have to be maintained at 20°C while the cameras are operated at -80°C. Both, mirrors and CCDs have to be kept within tight limits. The CCD cooling system consists of passive thermal control components only: two radiators, variable conductance heat pipes (VCHP) and two special thermal storage units. The orbit scenario imposes severe challenges on the thermal control system and also on the attitude control system.

Worldwide, 1.2 million people die in road crashes yearly; 43,000 in Europe alone. This implies a cost to European society of approximately 160 billion euros, and takes up 10% of all healthcare resources. To reduce these rates, safety technologies have been developed which help to minimize the severity of injuries to vehicle occupants. However, studies have shown that most deaths due to road accidents occur in the time between the accident and the arrival of medical care. Therefore, a fast and efficient rescue operation would significantly increase the injured person's probability of survival. The aim of this project was to define the On-Board Unit (OBU) hardware and software installed in all modern vehicles which could request medical and technical support after a road accident. This device, based on the information from the vehicle sensors, automatically decides whether the car has suffered a road accident or not, the severity of the accident and the kind of accident (impact area).

The permanently increasing number of convenience and safety functions leads to higher complexity of in-car electronics and the rapidly growing amount of sensors, actuators and electronic control units places higher demands on high- speed data communication protocols. Safety-critical systems need deterministic protocols with fault-tolerant behavior. The need for on-board diagnosis calls for flexible use of bandwidth and an ever-increasing number of functions necessitates a flexible means of extending the system. None of the communication solutions available on the market until now (like CAN or TTP) have been able to fulfill all these demands. To solve these problems, BMW together with several semiconductor companies has developed a new protocol for safety-critical applications in automotive vehicles.

A “next generation” condensing heat exchanger for space systems has to satisfy demanding operational requirements under variable thermal and moisture loads and reduced gravity conditions. Mathematical models described here are used to investigate transient behavior of wetting and de-wetting dynamics in the binary porous system of porous tubes and porous cold plate. The model is based on the Richard's equation simplified for the zero-gravity conditions. The half-saturation distance or the zone of influence of the porous annular suction tubes on the cold-plate porous material will be in the range of 1 to 10 cm for the time scales ranging from 100 to 10,000 seconds and moisture diffusivity in the range of D = 10-4 to 10-6 m2/s.

The preceeding presentation explained the compositions of the zinc alloys. Now we can examine the properties that these combinations give rise to. In doing so, we must first understand the importance of each property to the design of any component. One must consider each of the properties individually so as to gain an understanding of cumulative effects and the relative importance of each property to the final application. The most effective way to attack the problem is to analyze the part's requirements with respect to applied stresses, environmental and operating conditions and economic constraints. While there are many mathematical models available to enable numerical analysis of property evaluations, they all rely on the same basic principle: The Total Systems Approach. That is, the consideration of all aspects of design. Properties evaluation is a critical step in this process.

Zero-offset in transducer output during airbag noise testing is often observed, but mostly ignored due to the lack of understanding of its causes and implications. In the field of high-g acceleration measurement, this phenomenon is well documented, and is referred to as zeroshift. Zero-offset occurs when a component in the measurement chain is exposed to some unexpected inputs which the component has not been designed to handle. These unexpected inputs can be mechanical, electrical, or optical. How the transducer reacts to such inputs and the amount of zero-offset produced depends on the sensing mechanism, material used, and the design of the component itself. This paper explores the causes of zero-offset from a general perspective, covering the entire measurement chain. Although much of the information and discussions are based on data obtained from acceleration measurement systems, the findings are applicable to other transducer types, such as pressure and acoustic measurements.

As the demand for more complex system development and the ever-increasing requirement for improvement in software productivity, the need for graphical programming or Zero-Hand Coding for automatic generation of controller software becomes highly desirable. The graphical programming must not be limited to the algorithm development which consists of the application modules but must be extended to the microcontroller platform, which include the middleware (i.e. operating system, I/O device drivers) and hardware. Automatic code generation is very important for programming the complex microcontroller internal parameters and registers. The combined software tool chain is to generate the final target specific executable code. This approach is very beneficial for system development, reduction of the development cycle and bridges the gap between control and software engineers reducing time, effort and cost of the production software.

This paper describes the development of a set of algorithms that find the takeoff gross weight of an aircraft for given vehicle and engine characteristics, and mission requirements. A major objective was to find the most elementary set that would still yield useful answers. The result was a set that could be encoded on an inexpensive programmable pocket calculator with only 24 lines of code. Results are compared with actual characteristics of an executive jet and its derivative versions.