Search Results

This study outlines an approach for speeding up the simulation of the dynamic response of vehicle models that include hysteretic nonlinear tire components. The method proposed replaces the hysteretic nonlinear tire model with a surrogate model that emulates the dynamic response of the actual tire. The approach is demonstrated via a dynamic simulation of a quarter vehicle model. In the proposed methodology, training information generated with a reduced number of harmonic excitations is used to construct the tire hysteretic force emulator using a Neural Network (NN) element. The proposed approach has two stages: a learning stage, followed by an embedding of the learned model into the quarter car model. The learning related main challenge stems from the attempt to capture with the NN element the behavior of a hysteretic element whose response depends on its loading history.

The Service Bay Diagnostic System (SBDS) will be designed to assist the dealership technician in diagnosing and repairing Ford Motor vehicles. The system hardware will be configured around a Service Bay Computer with mass storage capability and auxiliary service equipment. Major system features include: guided service writer/customer interaction, interactive vehicle diagnostics, information management. capabilities, and an additional aid to identifying intermittent failures through the use of a portable over-the-road data acquisition device. In order to assist the technician in properly diagnosing the causal factor, the Service Bay Computer System will also be enhanced through the use of an expert system knowledge base.

When considering ride comfort and precision there are lots of components in the vehicle suspensions that have influence in this behavior and some ride occurrences (mainly higher frequencies) are rubber bushing responsibility but due their compliance, other vehicle attributes, steering and handling, can be affected. So the correct components tuning can maintain or improve vehicle attributes to address desired brand DNA and vehicle its specific needs. These studies were done considering the elastokinematics of front axle only due need of improve its comfort concerning higher frequencies (impacts and harshness). In addiction, correlation between subjective evaluation and objective data acquisition/post processing is desirable to optimize development time. Based in subjective directional, the activities time was reduced and final configuration reached faster.

This paper proposes a novel algorithm. This algorithm is called Self-Organizing Fuzzy Neural Network (SOFNN). SOFNN revolutionizes how researchers apply control theories, image/signal processing on control systems and other applications. In general, SOFNN is an identification technique that automatically initiates, builds and fine-tunes the required network parameters. SOFNN evaluates required structures without predefined parameters or expressions regarding systems. SOFNN sets out to learn and configure a system's characteristics. Self-constructing and self-tuning features enable SOFNN to handle complex, non-linear, and time-varying systems with higher accuracy, making systems identification easier. SOFNN constructs and fine-tunes the system parameter through two phases. The two phases are the construction and the parameter-tuning phase. The two phases run concurrently allowing SOFNN to identify systems on-line.

Neural networks are useful tools for optimization studies since they are very fast, so that while capturing the accuracy of multi-dimensional CFD calculations or experimental data, they can be run numerous times as required by many optimization techniques. This paper describes how a set of neural networks trained on a multi-dimensional CFD code to predict pressure, temperature, heat flux, torque and emissions, have been used by a genetic algorithm in combination with a hill-climbing type algorithm to optimize operating parameters of a diesel engine over the entire speed-torque map of the engine. The optimized parameters are mass of fuel injected per cycle, shape of the injection profile for dual split injection, start of injection, EGR level and boost pressure. These have been optimized for minimum emissions. Another set of neural networks have been trained to predict the optimized parameters, based on the speed-torque point of the engine.

Neural networks have been used for engine computations in the recent past. One reason for using neural networks is to capture the accuracy of multi-dimensional CFD calculations or experimental data while saving computational time, so that system simulations can be performed within a reasonable time frame. This paper describes three methods to improve upon neural network predictions. Improvement is demonstrated for in-cylinder pressure predictions in particular. The first method incorporates a physical combustion model within the transfer function of the neural network, so that the network predictions incorporate physical relationships as well as mathematical models to fit the data. The second method shows how partitioning the data into different regimes based on different physical processes, and training different networks for different regimes, improves the accuracy of predictions.

A flexible, easily configuration data acquisition system was designed and built for detailed studies of the steady state and dynamic properties of lean NOx traps for an engine dynamometer environment. The system is based on the industry standard VXI backplane. The overriding design philosophy was to design and develop a data acquisition system that was user friendly and could be operated easily by engine laboratory technicians, as well as by test engineers. The primary requirements guiding the design were the following: (1) the ability easily to configure, save, recall, modify, and print test configurations. (2) The ability to configure the gain, channel name, and engineering units for each analog channel. (3) The ability to trigger from one analog input channel. (4) A provision for numeric auto-incrementing of data file names. (5) The ability to save data in Excel™ compatible ASCII format. (6) The utilization of off-the shelf VXI hardware.

An experimental study of the performance of a reverse tumble, DISI engine is reported. Specific fuel consumption and engine-out emissions have been investigated for both homogeneous and stratified modes of fuel injection. Trends in performance with varying AFR, EGR, spark and injection timings have been explored. It is shown that neural networks can be trained to describe these trends accurately for even the most complex case of stratified charge operation with exhaust gas recirculation.

A number of advancements have been made in the coulometric sulfur trace instrumental technique for the real-time measurement of engine oil economy. These advancements include modification of the coulometric cell to improve reliability and reproducibility. The instrument has been interfaced with a microcomputer for instrument control as well as data acquisition, storage, and analysis. Studies were undertaken which demonstrate sufficient sensitivity and linearity for determination of engine oil economy at levels better than 10,000 miles/quart. Applications to steady-state engine oil consumption mapping and to instantaneous oil consumption during transient engine cycling are described. These instruments are being produced by an outside supplier for use in various company locations in both the engine production and engine research environments.

This paper is the second in the series of documents designed to record the progress of a series of SAE documents - SAE J2836™, J2847, J2931, & J2953 - within the Plug-In Electric Vehicle (PEV) Communication Task Force. This follows the initial paper number 2010-01-0837, and continues with the test and modeling of the various PLC types for utility programs described in J2836/1™ & J2847/1. This also extends the communication to an off-board charger, described in J2836/2™ & J2847/2 and includes reverse energy flow described in J2836/3™ and J2847/3. The initial versions of J2836/1™ and J2847/1 were published early 2010. J2847/1 has now been re-opened to include updates from comments from the National Institute of Standards Technology (NIST) Smart Grid Interoperability Panel (SGIP), Smart Grid Architectural Committee (SGAC) and Cyber Security Working Group committee (SCWG).

This paper describes the design and development of the Ford Driving Simulator. The simulator is a fixed-base device which provides real-time, interactive feedback to the driver through a combination of visual, auditory and tactile cues. The system is comprised of a modular buck, 150° field-of-view visual scene, a steering torque controller, high fidelity dynamics models, and an interactive experimenters station. Data acquisition systems have been developed to capture a broad spectrum of driver performance metrics.

During the fall of 1992, the Michigan Roadside Study was conducted. During this study IM240 tests were conducted on vehicles that had also been emissions tested during on-road operation via two remote sensors that were separated by 100 feet. The use of two remote sensors provided an indication of the short-term real-world emissions variability of a large number of on-road vehicles. This data was used to determine the frequency of flippers, i.e. vehicles that are sometimes high emitters (>4% CO) and at other times low emitters (<2% CO). The data show that the flipper frequency increases for older model year vehicles. Also, the correlations between remote sensor readings of emissions concentrations and IM240 mass emissions rates were determined. The data show that the correlation between remote sensing and IM240 improves with increasing numbers of remote sensing readings. For three remote sensor readings, CO correlates with an r2 of 0.69 and HC correlates with an r2 of 0.54

A highly adaptable fatigue testing computer system is presented for controlling single or multichannel test machines. The system imposes most common varieties of waveforms and also provides time synchronization between channels, such as in the case of variable amplitude biaxial load histories, and monitors various feedback signals for both data acquisition and alarm purposes. The program operates in a real-time Unix system as a separate stand-alone process. Communication with other users or the operator is done only through a reserved common block of shared memory. This feature allows control and monitoring of all tests over the computer network. A user can simply login remotely and check the test or start a data acquisition task from any workstation in the company, and then take the data files and analyze them on other computers. This paper describes the operation of the software, the methodology behind the hardware selection and the software structure.

A robotic sound intensity data acquisition system has been developed to overcome the problems of manual data acquistion. The system has the following performance: Accurate and repeatable positioning of the intensity probe Regular five and six sided rectangular measurement surfaces Irregular measurement surfaces around the complete powerplant, or individual components Continuous scanning or fixed point measurements Probe motion, data measurement, and storage of acoustic and geometric data integrated within the acquisition system database.

Automotive product development requires higher degree of quality upfront engineering, faster CAE turn-around, and integration with other functional requirements. Prediction of potential durability concerns using analytical methods for sheet metal structures subjected to road loads and other customer uses has become very important. A process has been developed to provide design direction based upon peak loads, simultaneous peak loads, and vehicle program analytical or measured loads. It identifies critical loads at each input location and load sets for multiple input locations, filters load time histories, selects critical areas and analyzes for fatigue life. Several case studies have been completed. The results show that the variations are consistent with the accuracies in finite element analysis, road load data acquisition, and fatigue calculation methods.

A new estimator for IC engine A/F ratio is described. A/F ratio is important for engine operation since it determines the quantities of engine emissions, such as HC, CO, NOx, the conversion efficiency of catalyst systems, and the engine combustion stability. The A/F ratio estimator described in this paper is based on a fundamental metric that relies on inducing and detecting crankshaft speed fluctuations caused by modulating the engine's fuel injection pulse widths. Fuel pulse width modulation varies the instantaneous combustion A/F ratio crankshaft velocity. Synchronous measurement of crankshaft velocity provides a metric that, when used with other engine state variables as inputs to a conventional neural network, can accurately estimate A/F ratio. The estimator provides A/F information when a physical sensor is not available.

This case study presents a new automated production noise test for power steering pumps. The test included adaptive noise cancellation, and a neural network implementation. The result mapped the pump acceleration signature into an objective repeatable noise metric. The test algorithm was a distributed DSP architecture designed for real-time measurement and decision processing. It was implemented with no increase in test cycle time. It accomplished the correlation of in-vehicle power steering pump noise to it's vibration characteristics, and retrofitting of accelerometers in place of microphones for acceptance testing.

A hybrid approach to predict road-induced loads in vehicle structures is presented. The technique involves full vehicle dynamic simulation using measured wheel forces, absolute wheel vertical displacements, and steering angle as input. The wheel vertical displacement is derived from the measured wheel acceleration. This approach avoids the use of tire-road interface modeling. It also improves the conventional loads measuring process with minimum instrumentation and data acquisition. Existing load data from a test vehicle is used to validate this approach. Computed component loads show good agreement with measurements.

A procedure has been developed to build system level predictive models that incorporate physical laws as well as information derived from experimental data. In particular a soot model was developed, trained and tested using experimental data. It was seen that the model could fit available experimental data given sufficient training time. Future accuracy on data points not encountered during training was estimated and seen to be good. The approach relies on the physical phenomena predicted by an existing system level phenomenological soot model coupled with ‘weights’ which use experimental data to adjust the predicted physical sub-model parameters to fit the data. This approach has developed from attempts at incorporating physical phenomena into neural networks for predicting emissions. Model training uses neural network training concepts.

A microcomputer-based, multichannel data acquisition system has been developed to acquire high frequency transient information typified by, but not limited to, automotive vehicle crash test applications. The system, which has been designed to be mounted on the test vehicle during a vehicle crash, will accommodate up to 240 channels. Each channel is comprised of a stand-alone microcomputer, memory for data storage, signal conditioning for piezoresistive transducers, automatic calibration and zero offsets, and programmable gain amplifier. The microcomputer is based upon a Motorola 6801/68701 microcomputer. The paper describes the design, development, and data processing characteristics of the prototype system.