Analysis of Autopilot Disengagements Occurring During Autonomous Vehicle Testing

In present-day highly-automated vehicles, there are occasions when the driving system disengages and the human driver is required to take-over. This is of great importance to a vehicle's safety and ride comfort. In the U.S state of California, the Autonomous Vehicle Testing Regulations require every manufacturer testing autonomous vehicles on public roads to submit an annual report summarizing the disengagements of the technology experienced during testing. On 1 January 2016,seven manufacturers submitted their first disengagement reports:Bosch, Delphi, Google, Nissan, Mercedes-Benz, Volkswagen, and Tesla Motors. This work analyses the data from these disengagement reports with the aim of gaining abetter understanding of the situations in which a driver is required to takeover, as this is potentially useful in improving the Society of Automotive Engineers(SAE) Level 2 and Level 3 automation technologies.Disengagement events from testing are classified into different groups based on attributes and the causes of disengagement are investigated and compared in detail. The mechanisms and time taken for take-over transition occurred in disengagements are studied. Finally, recommendations for OEMs, manufacturers, and government organizations are also discussed.

Experimental analysis of the slip sinkage effect based on real vehicle test

To improve the semi-empirical model, the slip sinkage effect is analyzed based on the real vehicle test. A dynamic testing system is used to gain the dynamic responses of wheel-soil interactions, The Gauss-Newton algorithm is adopted to estimate the undetermined parameters involved in the slip sinkage models. Wong＇s original model is compared with three typical slip sinkage models on the prediction performance of a drawbar pull. The maximum error rate, root mean squared error and correlation coefficient are utilized to evaluate the performance. The results indicate that the slip sinkage models outperform Wong＇s model and greatly improve the prediction accuracy. Lyasko＇s model is confirmed as an outstanding one for its comprehensive performance. Hence, the existence of the slip sinkage effect is validated. Lyasko＇s model is selected as an optimal one for the practical evaluation of military vehicle trafficability.

Fuzzy Sliding Mode Control for the Vehicle Height and Leveling Adjustment System of an Electronic Air Suspension

The accurate control for the vehicle height and leveling adjustment system of an electronic air suspension(EAS) still is a challenging problem that has not been effectively solved in prior researches. This paper proposes a new adaptive controller to control the vehicle height and to adjust the roll and pitch angles of the vehicle body(leveling control) during the vehicle height adjustment procedures by an EAS system. A nonlinear mechanism model of the full?car vehicle height adjustment system is established to reflect the system dynamic behaviors and to derive the system optimal control law. To deal with the nonlinear characters in the vehicle height and leveling adjustment processes, the nonlinear system model is globally linearized through the state feedback method. On this basis, a fuzzy sliding mode controller(FSMC) is designed to improve the control accuracy of the vehicle height adjustment and to reduce the peak values of the roll and pitch angles of the vehicle body. To verify the effectiveness of the proposed control method more accurately, the full?car EAS system model programmed using AMESim is also given. Then, the co?simulation study of the FSMC performance can be conducted. Finally, actual vehicle tests are performed with a city bus, and the test results illustrate that the vehicle height adjustment performance is effectively guaranteed by the FSMC, and the peak values of the roll and pitch angles of the vehicle body during the vehicle height adjustment procedures are also reduced significantly. This research proposes an effective control methodology for the vehicle height and leveling adjustment system of an EAS, which provides a favorable control performance for the system.

Hydrogen Shows Promise in Vehicle Tests

Car fleets escorted by police along Chang’an Avenue are not a rare sight for Beijing residents, who are accustomed to the practice when leaders of foreign states visit the capital of China.

Study of Vehicle Exhaust Variation with Test Modes

Nowadays harmful gas in vehicle exhaust has pollute d air heavily. To prevent the environment from polluting, the request of emissions control legislation becomes more stringent. New legislation prescribes not only the emissions limitation of vehicles, but also testing instruments and methods. Test car must be operated on the chassis dynamometer and data must be collect ed and analyzed with prescriptive exhaust analysis system as well. The mass of harmful exhaust gas, containing the concentration and volume of emis sion, which is independent from the model of automobile and engine, can be used as a criterion to evaluate the pollution of an automobile. Constant Volume Sampl e System (CVS) is used to measure vehicle emissions, but it is too expensive to apply extensively. The Vehicle Mass Analysis System(Vmas), a new vehicle exhaust mass analysis system produced in USA late 1990s,is used to test and analyze veh icle exhaust. As a test instrument, it has the virtue of cheapness and easy mana geability. In this paper, Vmas is used to measure the emissions of a light truck CA1020F. A ccording to 15 running modes of Vehicle Exhaust Legislation, the test car is ope rated on the chassis dynamometer and data are collected and analyzed with Vmas. The test results show that it is viable to measure and evaluate automobile emiss ion with Vmas. Most of HC exhaust is produced when the car is decelerating. The major factor that influences the mass of HC emission is the sudden decrease of e ngine load causing incomplete combustion in decelerating mode. Test results indi cate CO and NOx are mainly produced in the process of increasing load. The forme r reason is incomplete combustion and the latter is high burning temperature cau sed by the increasing load. The methods of reducing automobile emission are also discussed in this paper.

Theory of Vehicle's Durability Correlation Test

The theory of vehicle's correlation tests was discussed and the correlation test mathematical model was created. According to the damage theory, the correlation equation was brought up. The method to solove it was given. The text gives the theory basis of vehicle's correlation tests. The results can be used to give the vehicle's test designation and to make the test to fit the actual environments. Besides that, the methods give acceleration and strengthen effects.

Research on Modal Test Technology of LM-6A Solid-Liquid Strap-On Launch Vehicle

The LM-6 A new generation solid-liquid strap-on launch vehicle has the structural dynamic characteristics of lower frequencies,denser modes and coupling modes in longitudinal,bending and torsion modal space.During the development phase of LM-6 A,modal tests of partial stacks and the full vehicle were designed to obtain the structural dynamic properties.The structural dynamic models using the finite element method(FEM)have been verified and calibrated based on the modal test data.This paper describes the pre-test predictions and test execution,and details the comparison between the pre-test predictions and the test data.The successful maiden flight of LM-6 A further confirmed the effectiveness of structural dynamic modeling and modal test for LM-6 A.

Application of clutch to clutch gear shift technology for a new automatic transmission

For the purpose of engineering development for a new 8-step speed automatic transmission,a simplified dynamic model for this gearbox was established and key parameters which affected the shift quality were analyzed.Aiming at four different shift types,the ideal characteristics of shift clutch and engine control were set up.By using torque estimation method,PI slip control algorithm and engine coordinated control principle,the control model and transmission controller were well developed for three shift phases which included rapid-fill phase,torque phase and inertia phase.The testing environment on the rig and prototype vehicle level was built and the testing results obtained in ultimate condition could verify the accuracy and feasibility of this shift control strategy.The peak jerk during shift process was controlled within ±2 g/s where the smooth gearshift was obtained.The development proposal and algorithm have a high value for engineering application.

Method to Monitor Helicopter Rotor Fault Using Diagnostic Matrix

An efficient method for helicopter rotor fault diagnosis was proposed using diagnostic matrix and its software framework was developed.The diagnostic matrix was extracted by the way of simulating the quantitative models of damaged and undamaged rotor systems.The conflicts were recognized by the virtual symptom sensor based on the dependencies between symptoms and faults.Diagnostic candidates were generated by the Boolean hitting algorithm.The proposed method provides multi-fault real-time diagnostic capability and it is suitable for the synthetic health monitoring of complex systems.

Analysis on the gear-shifting schedule of hydraulic automatic transmission

With an advanced foreign hydraulic automatic transmission as the objective,an analytical method for the gear-shifting schedule is proposed.First the demanded maximum gradient of test is estimated.Then a test scheme and analytical procedure is formulated by initial test and hypothetical shift parameters.Finally through gear-shifting tests under different road conditions,load,accelerator pedal position limitation,throttle opening and output shaft speed are found to be the gear-shifting parameters.Under a common road condition,the gear-shifting schedule is a double-parameter schedule.Based on the driver＇s demands on braking and dynamic performance,different shift schedules are made under downhill,uphill and quick releasing acceleration pedal conditions.The operation criteria of down-shift schedule on abrupt grade are proposed.

SceGAN:A method for generating autonomous vehicle cut-in scenarios on highways based on deep learning

With the increasing level of automation of autonomous vehicles,it is important to conduct comprehensive and extensive testing before releasing autonomous vehicles into the market.Traditional public road and closed-field testing failed to meet the requirements of high testing efficiency and scenario coverage.Therefore,scenario-based autonomous vehicle simulation testing has emerged.Many scenarios form the basis of simulation testing.Generating additional scenarios from an existing scenario library is a significant problem.Taking the scenarios of a proceeding vehicle cutting into an adjacent lane on highways as an example,based on an autoencoder and a generative adversarial network(GAN),a method that combines Transformer to capture the features of a long-time series,called SceGAN,is proposed to model and generate scenarios of autonomous vehicles on highways.An evaluation system is established to analyze the reliability of SceGAN using discriminative and predictive scores and further evaluate the effect of scenario generation in terms of similarity and coverage.Experiments showed that compared with TimeGAN and AEGAN,SceGAN is superior in data fidelity and availability,and their similarity increased by 27.22%and 21.39%,respectively.The coverage increased from 79.84%to 93.98%as generated scenarios increased from 2,547 to 50,000,indicating that the proposed method has a strong generalization capability for generating multiple trajectories,providing a basis for generating test scenarios and promoting autonomous vehicle testing.

State Estimation of Vehicle’s Dynamic Stability Based on the Nonlinear Kalman Filter

An accurate estimation of a vehicle’s state of motion is the basis of dynamic stability control.Two different nonlinear Kalman filters are adopted for the estimation of the vehicle’s lateral/rollover stability state.First,the overall structure of the state estimation with four inputs and four outputs is introduced.After determining tire-cornering stiffness using a recursive leastsquares(RLS)method,the equations of state and of observation for the nonlinear Kalman filter are established based on a vehicle model with four degrees of freedom including planar and rollover dynamics.Then,the specific steps of real-time state estimation using the extended Kalman filter(EKF)and unscented Kalman filter(UKF)are both given.In a co-simulation,we find that the RLS algorithm estimates tire-cornering stiffness accurately and quickly,and the UKF improves the effect of state estimation compared with EKF.In addition,the UKF is verified against data from vehicle tests.The results show the proposed method is reliable and practical in estimating vehicle states.

Real drive cycles analysis by ordered power methodology applied to fuel consumption,C0_(2),NO_(x) and PM emissions estimation

In this work three fuel consumption and exhaust emission models,ADVISOR,VT-MICRO and the European Environmental Agency Emission factors,have been used to obtain fuel consumption(FC)and exhaust emissions.These models have been used at micro-scale,using the two signal treatment methods presented.The manuscript presents:1)a methodology to collect data in real urban driving cycles,2)an estimation of FC and tailpipe emissions using some available models in literature,and 3)a novel analysis of the results based on delivered wheel power.The results include Fuel Consumption(FC),CO_(2),NO_(x) and PM_(10) emissions,which are derived from the three simulators.In the first part of the paper we present a new procedure for incomplete drive cycle data treatment,which is necessary for real drive cycle acquisition in high density cities.Then the models are used to obtain second by second FC and exhaust emissions.Finally,a new methodology named Cycle Analysis by Ordered Power(CAbOP)is presented and used to compare the results.This method consists in the re-ordering of time dependant data,considering the wheel mechanical power domain instead of the standard time domain.This new strategy allows the 5 situations in drive cycles to be clearly visualized:hard breaking zone,slowdowns,idle or stop zone,sustained speed zone and acceleration zone.The complete methodology is applied in two real drive cycles surveyed in Barcelona(Spain)and the results are compared with a standardized WLTC urban cycle.