Road Test Research Articles

Real-Time Co-optimization of Gear Shifting and Engine Torque for Predictive Cruise Control of Heavy-Duty Trucks

Fuel consumption is one of the main concerns for heavy-duty trucks. Predictive cruise control (PCC) provides an intriguing opportunity to reduce fuel consumption by using the upcoming road information. In this study, a real-time implementable PCC, which simultaneously optimizes engine torque and gear shifting, is proposed for heavy-duty trucks. To minimize fuel consumption, the problem of the PCC is formulated as a nonlinear model predictive control (MPC), in which the upcoming road elevation information is used. Finding the solution of the nonlinear MPC is time consuming; thus, a real-time implementable solver is developed based on Pontryagin’s maximum principle and indirect shooting method. Dynamic programming (DP) algorithm, as a global optimization algorithm, is used as a performance benchmark for the proposed solver. Simulation, hardware-in-the-loop and real-truck experiments are conducted to verify the performance of the proposed controller. The results demonstrate that the MPC-based solution performs nearly as well as the DP-based solution, with less than 1% deviation for testing roads. Moreover, the proposed co-optimization controller is implementable in a real-truck, and the proposed MPC-based PCC algorithm achieves a fuel-saving rate of 7.9% without compromising the truck’s travel time.

An accurate trajectory tracking method for low-speed unmanned vehicles based on model predictive control

Trajectory tracking on a low-speed vehicle using the model predictive control (MPC) algorithm usually assumes a simple road terrain. This assumption does not correspond to the actual road situation, leading to low tracking accuracy. Therefore, a trajectory tracking method considering road curvature based on MPC is proposed in this paper. In this method, the controller can automatically switch between MPC types. Linear model predictive control (LMPC) is selected for small road curvatures, while nonlinear model predictive control (NMPC) is employed for large road curvatures. In addition, the NMPC algorithm in this work considers the effect of road curvature on tracking accuracy, making it suitable for tracking time-varying curvature roads. To verify the feasibility of the algorithm, simulation comparisons with the basic MPC model were carried out at different testing roads and vehicle longitudinal speeds. The results indicate that the method significantly improves trajectory tracking accuracy, all while ensuring real-time calculations. The intelligent switching capability of control models based on road curvature allows its application to track trajectories on arbitrarily complex roads.

Emission characteristic, spatial distribution, and health risk of polycyclic aromatic compounds (PAHs, NPAHs, and OPAHs) from light-duty gasoline and diesel vehicles based on on-road measurements

Vehicle exhaust is the primary source of polycyclic aromatic compounds (PACs). Real road tests using a portable vehicle measurement system on light-duty gasoline vehicles and light-duty diesel trucks were conducted to investigate gas- and particle-phase polycyclic aromatic hydrocarbons (PAHs), nitro-PAHs (NPAHs), and oxy-PAHs (OPAHs) in vehicle exhaust with different emission standards, fuel types, and altitudes. The results showed that with the tightening of emission standards, the overall emission factors (EFs) of PACs decreased. Compared with China V diesel vehicles, the emissions of PAHs, OPAHs, and NPAHs from China VI diesel vehicles were 75.1 %, 84.4 %, and 61.2 % lower, respectively. With a ∼100 m increase in altitude, the EFs of PAHs, OPAHs, and NPAHs of diesel vehicles increased 1.88, 1.92, and 1.59 times due to incomplete combustion. In addition, the EFs of PAHs and OPAHs in gasoline vehicles were lower than those in diesel vehicles. In contrast, the proportion of PAHs with highly toxic components, such as dibenzo[a,h]anthracene (DahA) and benzo[a]pyrene (BaP), and the EFs of gas-phase NPAHs in gasoline vehicles were higher than those in diesel vehicles. Furthermore, the emissions of 1,8-DNP from diesel vehicles cannot be disregarded. 1,8-DNP was the main gas-phase NPAHs emitted by China VI and China V diesel vehicles, accounting for 49.3 % and 26.0 %, respectively. Moreover, gas-phase PACs contributed more to the EFs than particle-phase PACs, whereas particle-phase PACs have greater toxic effects. Although the EFs of PAHs are more than 100 times those of NPAHs, the toxic equivalent concentrations (TEQBaP) of PAHs in diesel and gasoline vehicles were approximately 6.5 times and 35 times those of NPAHs. The spatial distribution characteristics revealed that PACs emissions were mainly concentrated in urban areas and highways, and the differences in the toxicity of PACs emissions between different cities depended on the proportion of diesel vehicles. The average TEQBaP of PAHs and NPAHs in Haidong, Haibei, Huangnan, Hainan, Guoluo, and Yushu was 8.42 μg/m3 and 0.36 μg/m3, respectively, while those of Xining and Haixi were 0.24–0.29 μg/m3 and 0.09–0.108 μg/m3 higher, respectively. This study provides a comprehensive understanding of the emission characteristics, health risks, and spatial distribution of PACs from diesel and gasoline vehicle PACs in urban areas.

Design and Aerodynamic Characteristics Analysis of an Electric Racecar Body Based on CFD

This study focuses on the development of a body for an electric racecar, utilizing CAD software for the design. A simplified full-vehicle geometric model was constructed. Based on fundamental theories of computational fluid dynamics and using CAE software platforms, the shear stress transport (SST) k-ω physical model was chosen to establish a three-dimensional computational model of the racecar’s external flow field. Simulations were conducted to analyze the pressure, airflow streamlines, and velocity distribution around the body and its surrounding flow field, elucidating the impact of body shape structure on aerodynamic characteristics. Finally, a manufacturing process for the body was designed, and a prototype was produced and integrated into the complete vehicle for road testing. The results indicate that the designed electric racecar body maintained consistent airflow over its surface, meeting the basic requirements of aerodynamics.

Design and Application of Rich Bottom Layer Mixtures

This study aims to summarize design indexes for rich bottom layer mixtures (RBLMs) and design asphalt mixtures with superior performance to reduce reflective cracks and fatigue cracking resulting from semi-rigid base cracking in the asphalt layer. First, this study synthesizes the research results about RBLMs and introduces the related design indexes. After that, using ABAQUS 6.13, finite element analysis was performed to analyze the change in bottom stress when the base cracked and the interlayer bond weakened. SBS-modified asphalt mixtures (SMMs) and composite-modified asphalt mixtures (CMMs) were subsequently developed as RBLMs utilizing the Marshall test method and applied on a test road. The research findings reveal that design indexes for RBLMs include a volume of air voids (VVs) within the range of 2.0–3.0%, a maximum flexural strain exceeding 3500με, and a dynamic stability surpassing 1000 cycles/mm. Both mixtures satisfy the design requirements, with the CMM demonstrating superior performance and offering promising application prospects. Future research endeavors will concentrate on prolonged monitoring of the test section to authenticate the effectiveness of an RBL in practical road applications.

Semi-supervised vanishing point detection with contrastive learning

Vanishing point (VP) detection in road images plays an important role in driving scenes for advanced driver assistance systems (ADAS) and autonomous vehicles. Because it is still a challenging problem to obtain expensive annotated datasets in deep-learning-based supervised training, in this paper, we propose a semi-supervised VP detection method in road images. Our proposed model first extracts high-resolution VP heatmaps of the road images by fusing multi-level and global–local contrastive learning. Then, we apply the π model as a semi-supervised learning framework and feed the whole global network with a small number of training samples to detect VPs. In the experiment, we used Kong’s dataset for the unstructured road test and the KITTI-VP dataset for the structured road test. Compared with the state-of-the-art methods, our model achieved high accuracy and robustness in road VP detection.

Adaptation and Validation of Injection Rate Predictive Model for Solenoid Type Injectors with Different Nozzle Geometry

The present research analyses the injection rate of a direct rail injection diesel engine, focusing specifically on the influence of the nozzles and various operating conditions from real road tests on the rate of injection. A diesel injector test bench was used for feedback with real data from the test vehicle under real road conditions. An analysis of the behaviour of the injection rate was carried out using the zero-dimensional model. This model generated a predictive model that incorporated the five variables identified through a developed multivariate analysis of variance, showing a high correlation of dependence between variations in injection pressure, the diameter of the holes, and the number of holes with greater representativeness. The results obtained showed that the nozzle geometry and the physical properties of the fuel had a direct effect on the injection rate. This analysis enriches the understanding of fuel injection and its effects on diesel engine performance by providing an analysis of the system components that influence the injection rate and generating a simple tool to feed thermodynamic diagnostic models. The proposal model may be used as an input in thermodynamics predictive models and reduce the simulation load in computational fluid dynamics predictive models.

Strain-based running-reliability characterisation in time-domain for risk monitoring under various load conditions

This aim of this paper is to characterise the strain-based fatigue life data in time-domain using the newly modelled running-reliability technique that considers the load sequence effect. Current established conventional strain life models do not consider dependence for fatigue life of low or high amplitudes, on which with occur first in the load history. Finite element analysis is carried out to ensure the strain signals are captured at the most critical region during road test at various conditions. Fatigue life of 2.74 × 104 to 6.07 × 105 cycle/block with mean cycle to failure of 4.32 × 106 to 7.00 × 106 cycle/block is predicted based on the cycle sequence effect using cycle-counting method. The newly modelled running-reliability technique is formulated to extract the features of high amplitude excitation obtained from the strain signals for characterising the fatigue reliability features under load sequence effect. Hence, the reliability-hazard relationship for fatigue reliability characterisation of strain-based approach in time-domain using running-reliability technique.

Modelling CO2 Emissions from Vehicles Fuelled with Compressed Natural Gas Based on On-Road and Chassis Dynamometer Tests

In response to increasingly stringent global environmental policies, this study addresses the pressing need for accurate prediction models of CO2 emissions from vehicles powered by alternative fuels, such as compressed natural gas (CNG). Through experimentation and modelling, one of the pioneering CO2 emission models specifically designed for CNG-powered vehicles is presented. Using data from chassis dynamometer tests and road assessments conducted with a portable emission measurement system (PEMS), the study employs the XGBoost technique within the Optuna Python programming language framework. The validation of the models produced impressive results, with R2 values of 0.9 and 0.7 and RMSE values of 0.49 and 0.71 for chassis dynamometer and road test data, respectively. The robustness and precision of these models offer invaluable information to transportation decision-makers engaged in environmental analyses and policymaking for urban areas, facilitating informed strategies to mitigate vehicular emissions and foster sustainable transportation practices.

Effect of Shock Absorber Friction on Vehicle Vertical Dynamics

<div>In order to efficiently predict and investigate a vehicle’s vertical dynamics, it is necessary to consider the suspension component properties holistically. Although the effects of suspension stiffness and damping characteristics on vertical dynamics are widely understood, the impact of suspension friction in various driving scenarios has rarely been studied in both simulation and road tests for several decades.</div> <div>The present study addresses this issue by performing driving tests using a special device that allows a modification of the shock absorber or damper friction, and thus the suspension friction to be modified independently of other suspension parameters. Initially, its correct functioning is verified on a shock absorber test rig. A calibration and application routine is established in order to assign definite additional friction forces at high reproducibility levels.</div> <div>The device is equipped in a medium-class passenger vehicle, which is driven on various irregular road sections as well as over single obstacles. For all tested road sections, a linear decrease of ride comfort in terms of specific relevant vertical objective values is found by increasing the friction force. This emphasizes a definite link between suspension friction and vertical vehicle body vibration, resulting in a negative impact on vertical ride comfort. However, the longitudinal vehicle body vibration is not significantly affected.</div> <div>The relevance of friction in terms of transmitting the energy associated with road unevenness to the chassis in the frequency range of the chassis’ natural frequency is found to be remarkably high on smooth roads, and still considerably high on bumpy roads. The chassis and wheel resonance frequencies are significantly friction-dependent due to the damper’s slip or stick states. The results obtained from smooth road tests demonstrate the practical relevance of accurately considering friction for the given suspension type in terms of vertical ride comfort prediction.</div>

The energy-saving effect of early-stage autonomous vehicles: A case study and recommendations in a metropolitan area

This paper presents the energy-saving potential of early-stage autonomous vehicles (AVs) by analyzing empirical AV driving records from a public road test. Vehicle-specific power-based microscopic energy consumption models are used to estimate the energy efficiency of the tested AVs, including light-duty vehicles and heavy-duty trucks. Overall, AVs do not necessarily save energy, largely depending on driving scenarios, such as traffic conditions and road types. On expressways, autonomous driving (AD) has an insignificant energy-saving effect (within 10 %) due to similar characteristics to human driving (HD). On urban roads with lower traffic speeds and more interventions from surrounding traffic, AD performs higher energy efficiency due to its capability to accelerate more smoothly by up to 60 %; however, frequent human disengagement generates additional energy consumption of 8%–40 %, especially on urban arterials. We further assessed the marginal effect of trip average speed and AD ratio on energy efficiency, illustrating varied relationships for different driving scenarios and power types. The study suggests a more moderate AD acceleration to elevate the energy-saving effect. Nevertheless, applying AD can hardly reduce the energy of early-stage AVs; instead, control methods to maintain an appropriate speed of the mixed traffic flow can help decrease the total energy consumption.

Have climate change and warmer winters altered freeze-thaw patterns?

Although an increase in the number of freeze–thaw (F/T) events are known to cause increased damage to pavement systems, further research is required to understand the impact of climate change on F/T cycles. This study aimed to determine the frequency of F/T cycles, and hence damage events, at typical air temperature sensor levels, in various layers of pavement, and at critical subgrade locations. In response to global warming, the study considered rigorously computed F/T events rooted in Minnesota climate data between 1901 and 2020. Cooperative Observer Network (COOP) and Modern-Era Retrospective Analysis Version 2.0 (MERRA-2) data were employed to examine subsurface warming conditions. The number of F/T cycles at the air temperature sensor level and individual pavement layers was determined by utilizing historical data from the COOP weather stations and Minnesota Road Test Facility (MnROAD). The evaluation was conducted on both a monthly and yearly basis. Results indicated that starting from 1981, Minnesota winters have become warmer by 0.1 to 1.2 °C in the daytime and by 0.3 to 1.5 °C during nighttime. The warmer winter weather results in fewer F/T cycles at shallow pavement levels on an annual basis while remaining erratic at deeper sub-pavement locations. The reduction of F/T events at shallower pavement depths was substantial during the late fall and early spring months. While monthly composites of F/T events at air-temperature sensor levels also reflected a decrease in F/T cycles during late fall and spring, the annual frequency of F/T events was randomly distributed.

Study on fatigue performance of asphalt mixture in service life based on accelerated loading test

To further study the fatigue damage evolution law of asphalt mixture in service period, the accelerated pavement testing (APT) of graded gravel base asphalt pavement and in-situ pavement was conducted. The full-scale test road of graded gravel base asphalt pavement was built, and the mechanical response sensors and temperature sensors were embedded. The fatigue damage test of the full-scale test road was conducted at 15°C through the accelerated loading test (ALT) equipment. The mechanical responses under the repeated wheel loadings were captured by the strain sensor at the bottom of the asphalt layer, and the deflections of the pavement under different cycles of wheel load were tested by Falling Weight Deflectometer (FWD). The fatigue damage model of asphalt mixture was established, and the accuracy of the model was verified by the accelerated loading fatigue test of in-situ pavement. The results show that asphalt layer bears an alternating effect of tensile and compressive strain, and when the wheel load is reduced or the driving speed increases, the tensile strain at the bottom of the asphalt layer decreases. As an evaluation index of fatigue damage of asphalt layer, the residual modulus ratio can accurately characterize the cumulative degree of fatigue damage. The modulus of asphalt layer shows an inverse S-type curve with the repeated wheel loadings. Based on the quantitative analysis of the whole process of fatigue damage of asphalt layer, a nonlinear damage evolution model of modulus attenuation of asphalt mixture was established, which would provide a reference for the study of pavement service performance attenuation law in different climatic environments and structural compositions in the future.

Optimization of the installation techniques of distributed optic fiber sensors and measurement accuracy analysis for asphalt pavement based on finite element methods

The accurate measurement strain is critical to evaluating the structural integrity of asphalt pavements. The measurement strain error is caused by the encapsulation and installation of distributed optical fiber sensor (DOFS) for pavement monitoring. This study introduces an innovative approach to correct installation error. Firstly, a DOFS is embedded into asphalt pavement to simulate the mechanical response caused by traffic load. The optimal installation technology for each asphalt layer is proposed. In addition, the strain correction formula considering the pavement temperature is proposed for different asphalt layer. A rutting test is used to evaluate monitoring performance. Finally, a test road was built. A continuous temporal and spatial distribution strain of DOFS caused by traffic load is obtained. The results indicate that the average error rate of strain testing effectively decreases for each asphalt layer. In summary, this paper contributes to obtaining more accurate measurement strain of DOFS in pavement monitoring.

Study on physiological representation of passenger cognitive comfort: An example with overtaking scenarios

During autonomous driving, the subjective cognition of scenario risks may lead to cognitive discomfort of occupants. This is a new concept of ride comfort for vehicles, which is being included in the performance evaluation of autonomous vehicles. The observation methods, generation mechanisms, prediction methods, and optimization methods of this cognitive discomfort remain to be explored. In order to quantify the risk cognition of passengers and evaluate cognitive comfort, this paper proposes a physiological representation method using overtaking scenarios as an example. Firstly, physiological data of passengers are collected through road test, and 19 physiological indicators are extracted, including 6 galvanic skin response (GSR) indicators, 5 electrocardiogram indicators, and 8 eye movement indicators. Secondly, the physiological indicator set related to the subjective evaluation results of comfort is constructed through correlation analysis. Finally, a multiple linear regression model is established to represent the level of passenger discomfort, with physiological indicators as inputs and subjective comfort indicators as outputs. The indicators that contribute significantly to the output results in the model are the maximum amplitude of GSR, the rise time of GSR, the maximum pupil diameter, and the average fixation time. The validation results indicate that the overall accuracy of the model is 86.90 %.

Research on surface treatment technology for quickly improving the skid resistance of tunnel concrete pavement.

In order to solve the problem that the skid resistance of concrete pavement in tunnel deteriorates rapidly, which is easy to cause traffic accidents, the anti-skid rapid elevation technology of surface treatment is proposed. Wear tests were used to investigate the effects of concrete surface roughness, properties of modified emulsified asphalt binder and anti-skid fine aggregate type on long-term skid resistance of treated surfaces. The results show that the four coarsening methods of fine milling, milling, grooving and brooming can improve the skid resistance of concrete, and the skid resistance durability of fine milling and milling is better. The adhesive property of modified emulsified asphalt is the best when the content of water-based epoxy resin is 20%. In different aggregates, the anti-skid effect is better when silicon carbide is used as anti-skid aggregate and the particle size is 0.6mm:0.3mm = 2:3. The method of fine milling of concrete surface + spraying epoxy emulsified asphalt + spreading silicon carbide can effectively improve the anti-skid performance of the original concrete pavement, and the feasibility of the scheme is verified by the test road. The research results have a good reference value for improving the skid resistance of tunnel concrete pavement.

A Kalman Filtering Algorithm for Measurement Interruption Based on Polynomial Interpolation and Taylor Expansion.

Combined SINS/GPS navigation systems have been widely used. However, when the traditional combined SINS/GPS navigation system travels between tall buildings, in the shade of trees, or through tunnels, the GPS encounters frequent signal blocking, which leads to the interruption of GPS signals, and as a result, the combined SINS/GPS-based navigation method degenerates into a pure inertial guidance system, which will lead to the accumulation of navigation errors. In this paper, an adaptive Kalman filtering algorithm based on polynomial fitting and a Taylor expansion is proposed. Through the navigation information output from the inertial guidance system, the polynomial interpolation method is used to construct the velocity equation and position equation of the carrier, and then the Taylor expansion is used to construct the virtual measurement at the moment of the GPS signal interruption, which can make up for the impact of the lack of measurement information on the combined SINS/GPS navigation system when the GPS signal is interrupted. The results of computer simulation experiments and road measurement tests based on the loosely combined SINS/GPS navigation system show that when the carrier faces a GPS signal interruption situation, compared with a combined SINS/GPS navigation algorithm that does not take any rescue measures, our proposed combined SINS/GPS navigation algorithm possesses a higher accuracy in the attitude angle estimation, a higher accuracy in the velocity estimation, and a higher accuracy in the positional localization, and the system possesses higher stability.

Evaluation of the Effect of De-icing Chemicals on Performance of Airport Concrete Pavement under Freeze-Thaw Cycles

Weather conditions such as rainfall and freeze-thaw cycles affect the pavement performance of airports; therefore, methods such as using de-icing chemicals are considered in order to maintain the normal condition of the airport runway. In addition to the above factors, de-icing chemicals play an important role in pavement performance and the damage caused to it due to their chemical nature. Therefore, investigating the effect of de-icing chemicals and determining the appropriate material to maintain the airport's pavement is a priority for engineers. In this study, the effect of three de-icing chemicals, sodium chloride, potassium acetate, and ethylene glycol (at concentrations of 23.3%, 49%, and 69.07%, respectively) on skid resistance (The British Pendulum Test (BPT) and The Road Test Machine (RTM)) and the mechanical properties (Compressive Strength Test) of the concrete under Freezing and Thawing Cycle were investigated. The BPT test results showed that sodium chloride resulted in better skid resistance than other chemicals when the number of cycles is more than 100 cycles. This result was also obtained for all cycles in the RTM test. Also, ethylene glycol was not suitable for improving skid resistance based on BPT and RTM tests. Furthermore, the results of the compressive strength of concrete mixtures showed that the de-icing chemicals reduced the compressive strength of concrete mixtures. Based on all the results, sodium chloride had better results than other chemicals.

The design and development of an Integrated Propulsion System – Phase 3: the testing bench

The paper presents the testing bench dedicated to the integrated propulsion system (IPS). The IPS consists of a wheel that carries the electric machine and the suspension inside. Its behaviour needs to be tested while being used independently on a test bench and then on road tests. Because no already existing testing bench was available for this kind of testing, a dedicated testing bench was designed and developed. Its main role is to investigate the functional behaviour of the wheel, together with the integrated suspension system behaviour. The tests were made following different scenarios, where different tire dimensions, different environment properties, different road profiles etc. were reproduced.

G-DGANet: Gated deep graph attention network with reinforcement learning for solving traveling salesman problem

Combinatorial optimization problem (COP) is an NP-hard problem for which finding an optimal solution is difficult, especially as the problem size increases. The Traveling Salesman Problem (TSP), one of the COPs that can be formulated over a graph, is a well-researched area in operations research and computer science. Deep Reinforcement Learning (DRL) is now regarded as a promising approach for solving TSP and other NP hard problems. In this paper, we propose a novel Gated Deep Graph Attention Network (G-DGANet) which builds upon the existing Graph Neural Network (GNN) to solve TSP. The proposed G-DGANet uses gating mechanism between subsequent layers of the network to extract representations of nodes deeper in the network without loss in performance. G-DGANet also designs a novel aggregator to construct global graph embeddings from different embedding preferences. In addition, to effectively learn underlying structure of a graph, G-DGANet integrates node and edge information of the graph while updating node representations in the message passing mechanism. We used proximal policy optimization (PPO) to train G-DGANet on randomly generated instances. We conducted an experiment on randomly generated instances and on real-world road network data generated from digital maps to verify the performance of G-DGANet. The findings from experiments demonstrate that G-DGANet outperforms most traditional heuristics and existing DRL approaches, with high generalization abilities from random instance training to random instance testing and real-world road network instance testing.