Search Results

An obstacle recognition algorithm for the Pre-Crash Safety system has been newly developed with a stereo vision system and a millimeter wave radar with additional functions. This algorithm uses the merits of both the millimeter wave radar and the stereo vision system, and has two main features. One feature utilizes the merits of the stereo vision system detection with the detection results from the millimeter wave radar allowing for a more detailed horizontal position and width of the obstacle. This enables the equipment to operate at an earlier stage according to how well the relationship between the vehicle and the obstacle is understood. Another feature fuses detection from the millimeter wave radar and the stereo vision system. This system has succeeded in enhancing the detection performance of pedestrians who have been more difficult to detect than reflective objects such as cars.

The purpose of this study is to propose braking characteristics that are easy for drivers to handle in a system in which braking and driving operations are performed by hand. Genetic algorithm optimization of braking characteristics showed that the best deceleration tracking was achieved by an FG diagram with a logarithmic function shape. In contrast, the slope of the optimal FG diagram tended to decrease as the driver's proportional gain increased.

A driving simulator system for developing steering road feel has been developed. A new steering gear box or an electronic steering system is installed on the simulator and its road feel and control algorithm are developed according to the characteristics of any vehicle which has been programed into the engineering work-station. The vehicle model programed into the engineering work station runs according to the driver's operations, which are fed through the new steering system to be tested. The steer-restoring torque of the vehicle programed into the engineering work-station is produced by an actuator, and gives the impression through the new system of having been fed back from an actual road.

One of the most common applications of planetary (epi-cyclic) gear sets is found in automotive transmissions. A planetary gear set typically total torque applied to be shared by multiple planets making a higher power density possible. This advantage of the planetary gear sets relies heavily on the assumption that each pinion carries an equal share of the total torque applied. However, in production, gear manufacturing and assembly variations along with certain design parameters may prevent equal load sharing among the planets. Here, a generalized mathematical model of a single-stage planetary gear set having n planets is developed to predict load shared by each planet under quasi-static conditions. The model takes into account effects of two most common errors including pinion carrier errors and gear run-out errors. Results of an experimental test program are used to validate the predictions of the model. Generalized guidelines for equal load sharing are also presented.

Toyota Safety Sense is a safety system package developed to help drivers avoid accident types with a high frequency of occurrence. This paper deals with pre-collision system which forms the core of Toyota Safety Sense, especially Toyota Safety Sense P which uses a combined sensor configuration consisting of a monocular camera paired with millimeter wave radar, in order to achieve both high recognition performance and reliability. The use of a wide-angle monocular camera, millimeter wave radar integrated in the front grill emblem, and a collision determination algorithm for pedestrian targets enabled the development of a pre-collision system comprising detection capability of crossing pedestrians. Toyota has developed warning and pre-collision brake assist for driver to assist in avoiding a collision effectively; In addition, Pre-collision brake has achieved high level of performance for the drivers who cannot avoid a collision.

Many vehicles are currently equipped with active safety systems that can detect vulnerable road users like pedestrians and bicyclists, to mitigate associated conflicts with vehicles. With the advancements in technologies and algorithms, detailed motions of these targets, especially the limb motions, are being considered for improving the efficiency and reliability of object detection. Thus, it becomes important to understand these limb motions to support the design and evaluation of many vehicular safety systems. However in current literature, there is no agreement being reached on whether or not and how often these limbs move, especially at the most critical moments for potential crashes. In this study, a total of 832 pedestrian walking or cyclist biking cases were randomly selected from one large-scale naturalistic driving database containing 480,000 video segments with a total size of 94TB, and then the 832 video clips were analyzed focusing on their limb motions.

This paper outlines a newly developed method for predicting the coupled response of structures from their uncoupled forced responses without having to know the forces acting on such structures. It involves computing the forced response of originally uncoupled structures with several mass loadings at a potential coupling point. The response data obtained from such computations is then used to predict the coupled response. The theory for discrete linear systems is outlined in the paper and a numerical example is given to demonstrate the validity, advantages and limitations of the method. The method is primarily devised to obtain coupled response of linear dynamic systems from independent and uncoupled analytical simulations. Its application significantly decreases computation time by reducing the simulation model size and is excellent for “what if” scenarios where a large number of simulations would otherwise be necessary.

This paper presents a pioneer work and first time application of Multi Objective Robust Optimization to analytically improve Idle Shake Performance. The method is developed to obtain a robust design with multiple objectives under consideration along with managing material property variation. It was a Robust Optimization on top of Multi Objective Genetic Algorithm, Robu-MOGA. The design variables in the study included the nominal values and tolerances of Sound Transmission Loss property, and interior material Absorption property. The analytical objective was not only to minimize the peak airborne noise at each specified frequency, but also to reduce the total cost and the total mass of the materials. In the study, AutoSEA (statistical energy analysis) from ESI Software, Inc. was used as the solver. AutoSEA was integrated with iSIGHT from Engineous Software, Inc.

Hardware-in-the-Loop (HWIL) testing is a means for validating and verifying component designs in a system context. Most current HWIL work with electrochemical systems for automotive applications has focused on the pack level, providing valuable feedback to system designers. Further benefits are realized by implementing this concept earlier in the development process; applying test vectors to an individual cell, but attenuating the stimulus and feedback to pack levels. This paper reports on a cell-level HWIL system designed to evaluate electrochemical cells and associated subsystems for advanced hybrid-electric vehicles (HEVs). The architecture of the system is described along with an example of its application applied to a commercially available supercapacitor and a state-of-charge algorithm in an HEV-based configuration.

The value of model-based design has been attempted to be communicated for more than a decade. As methods and tools have appeared and disappeared from a series of different vendors it has become apparent that no single vendor has a solution that meets all users’ needs. Recently standards (UML, MDA, MOF, EMF, etc.) have become a dominant force and an alternative to vendor-specific languages and processes. Where these standards have succeeded and vendors have failed is in the realization that they do not provide the answer, but instead provide the foundation to develop the answer. It is in the utilization of these standards and their capability to be customized that companies have achieved success. Customization has occurred to fit organizations, processes, and architectures that leverage the value of model-driven design.

A rapid prototyping controller (RPC) based, full-authority, diesel control system is developed, implemented, tested and validated on FTP cycle. As rapid prototyping controller, dSPACE Autobox is coupled with a fast processor based slave for lower level I/O control and a collection of in-house designed interface cards for signal conditioning. The base software set implemented mimics the current production code for a production diesel engine. This is done to facilitate realistic and accurate comparison of production algorithms with new control algorithms to be added on future products. The engine is equipped with all the state-of-the art subsystems found in a modern diesel engine (common rail fuel injection, EGR, Turbocharger etc.).

A vehicle parking manoeuvre is characterized by low or zero speed, small turning radius and large yaw velocity of the steered wheels. To predict the forces and moments generated by a wheel under these conditions, the Pacejka Magic Formula model has been extended to incorporate the effect of spin (turn slip model) in the past years. The extensions have been further developed and incorporated in the MFTyre/MF-Swift 6.2 model. This paper describes the development of a method for the identification of the turn slip parameters. Based on the operating conditions of a typical parking manoeuvre, the dominant parameters of the turn slip model are firstly defined. At an indoor test facility, the response of a tyre under the identified operating conditions is measured. An algorithm is developed to identify the dominant turn slip parameters from the measured responses.

A few reports suggest differences in injury outcomes between cadaver tests and real-world accidents under almost similar conditions. This study hypothesized that muscle activity could primarily cause the differences, and then developed a human body finite element (FE) model with individual muscles. Each muscle was modeled as a hybrid model of bar elements with active properties and solid elements with passive properties. The model without muscle activation was firstly validated against five series of cadaver test data on impact responses in the anterior-posterior direction. The model with muscle activation levels estimated based on electromyography (EMG) data was secondly validated against four series of volunteer test data on bracing effects for stiffness and thickness of an upper arm muscle, and braced driver's responses under a static environment and a brake deceleration.

For the purpose of reducing fuel consumption, a hybrid heavy duty truck was considered. Generally, HV (Hybrid Vehicle)'s energy is regenerated from deceleration energy in urban area. Hybrid heavy duty truck's energy is regenerated from potential energy on highway. Under this circumstance, some portion of energy may not be accumulated, because capacity of HV battery is limited. In order to maximize accumulating energy in the next descent, HV battery's energy shall be adequately reduced beforehand. This can be achieved by optimizing motor assist torque considering road's altitude and gradient. In this paper, performance of the algorithm is discussed.

This paper gives an outline of the steering assistance system (hereinafter, SAS) and a description of its key technology: the lane recognition algorithm. To accommodate a variety of driving styles, the SAS is equipped with a lane keeping assistance mode (LKA mode) and a lane departure warning mode (LDW mode) that can be selectively set by the driver. The former mode works in combination with adaptive cruise control (ACC) and carries the advantage of relieving the driving load that is placed on the driver. The latter mode has the benefit of reducing the danger of lane departure accidents caused by the driver dozing off and taking his eyes off the road. The newly developed lane recognition ECU has a simple hardware set-up of two 32-bit microcomputers. The lane recognition algorithm was constructed on the basis of a logic process that analyzes pattern edge points and selects a set of edge points that most closely resemble lanemarks.

This paper covers the application of hybrid vibro-acoustic simulation methods to shorten the design cycle of power-plant components. A comparison is made between Frequency Response Function based and Modal based algorithms for the generation of a predictive powerplant assembly model. The effectiveness of design modifications is evaluated by loading the original and modified predictive models with experimentally identified excitation forces. The procedure is validated by correlation with experimental data.

A hardening depth evaluation technique using eddy current has been developed, which can be applied to a mass production line for destructive (cutting) inspections. Using this technique, changes in the hardness of the induction-hardened structure can be detected based on the changes in magnetic permeability. This technique reduces the thermal effect and improves measurement accuracy through a multi-frequency exciting method and a difference method algorithm.

Toyota Motor Corporation has already designed and developed vehicle brake control systems for relatively low speed off-road driving, such as Downhill Assist Control, Hill-start Assist Control and Active Traction Control. Though off-road utility is improved by virtue of these systems, in specific situations actual performance still depends on driving technique since the driver is required to control the accelerator pedal. Toyota has integrated these existing systems, and developed a new driving technology for off-road driving called “Crawl Control.” Crawl Control automatically modulates brake torque and drive torque to help keep the vehicle speed constant and slow. Unskilled drivers can thereby attain improved capabilities in places where high-level driving techniques are required. This system also reduces the effort required to control the accelerator and the brake pedal. This paper presents a new control algorithm for the realization of this Crawl Control system.

In recent years, GPFs (Gasoline particulate filters) have been installed in gasoline engines to comply with stricter environmental regulations in China and Europe. In particular, coated-GPFs having a catalytic purification function are required to have high conversion performances, high filter efficiencies in the sense of a high collection efficiency, and low pressure loss. It is not easy to design a filter that satisfies all these parameters. Experimental studies are being conducted, but it is costly to study in trial productions. In this technical paper, a GPF design optimization method will be proposed that combines multi-scale simulation, surrogate models by machine learning, and an optimization algorithm. By using this method, a GPF design that minimizes pressure loss while providing high conversion performance and particle collection rates that satisfy current regulations can be created.

Conventional CFD-based shape optimization technology that uses parametric shape modification and optimal solutions searching algorithms has the two problems: (1) outcome of optimized shapes depend on the selection of design parameters made by the designer, and (2) high computational costs. To resolve those problems, two innovative inverse analysis technologies based on the Adjoint Method were developed in previous study: surface geometry deformation sensitivity analysis to identify the locations to be modified, and topology optimization to generate an optimal shape for maximizing the cost function in the constrained design space. However, these technologies are only applicable to steady flows. Since most flows in a vehicle (such as engine in-cylinder flow) are transient, a practical technology for surface geometry sensitivity analysis has been developed based on the Transient Adjoint Method.