On-road Vehicles Research Articles

Prognostics of battery capacity based on charging data and data-driven methods for on-road vehicles

The large-scale application of lithium-ion batteries makes it urgent to accurately predict their capacity degradation so as to achieve timely maintenance and second-life utilization. For on-road electric vehicles (EVs), due to limitation of battery management system in measurement and computing power, it is still a tricky challenge to accurately predict the capacity of battery pack. To this end, a battery capacity prognostic method based on charging data and data-driven algorithms is proposed in this paper. First, battery capacity is calculated based on a variant of Ampere integral formula, and statistical values of the capacity during a month are regarded as labeled capacity to reduce errors. Then, statistical characteristics of battery charging data are extracted, and correlation analysis and feature selection are conducted to determine optimal feature sets. Moreover, a sequence-to-sequence (Seq2Seq) model is employed to predict future capacity trajectory, and two residual models based on Gaussian process regression (GPR) are proposed to compensate the prediction error caused by local capacity change. Finally, the data of 20 EVs operating about 29 months are used to verify the proposed methods. By using the first 3 months data as input, the remaining capacity sequence can be accurately predicted with error lower than 1.6%.

Deep learning for 6D pose estimation of objects — A case study for autonomous driving

Nowadays, the potential benefits and implementation of autonomous driving have attracted widespread attention from both industry and academia. This study will solve view-invariant object detection and semantic key-point pose assumptions from a single RGB image. A machine learning method for estimating the absolute pose of an on-road vehicle for autonomous driving from monocular vision alone without the help of additional sensors is a complex task. The main purpose of this work is to identify other vehicles on the road and estimate their exact angular position from a single image with improved accuracy. The focus of the study is to create a new algorithm by applying a potentially deep convoluted neural network followed by a repetitive neural structure for more accurate 6D pose inference. A 6D pose hypothesis is presented in this study, based on a deep hybrid architecture for individual vehicles of an end-to-end approach to a task consisting of a Convolutional Neural Network (CNN) and a Recurrent Neural Network (RNN). In this work, we will use a large-scale dataset consistent with the understanding of a 3D car instance called ApolloCar3D. The data set contains 5,277 real-life street scenes with examples of about 60K cars. By comparison, the ApolloCar3D is twenty times larger than the PASCAL3D+ and KITTI datasets. Ultimately, the idea is to efficiently eliminate motionless cars and predict the next pose given in the speed context, allowing a comprehensive evaluation, and passing the output through LSTM (long short-term memory) with an additional filter layer. The new filter added to the LSTM will efficiently filter and isolate stationary or parking vehicles and focus on on-road vehicles. Since the LSTM has a non-linear high-dimensional hidden memory state, it can preserve the past continuity of each generation’s data history and pay more attention to those road vehicles rather than parked or stationary vehicles to act accordingly. So for each new vehicle, the pose estimator classifier can use LSTM memory and compare the historical pose with the newly filtered data. The successful implementation of this innovative concept will lead to significant improvements in the real-life traffic situation in the field of computer vision and autonomous driving.

Direct Observation of HONO Emissions from Real-World Residential Natural Gas Heating in China.

Atmospheric nitrous acid (HONO) is an important precursor of atmospheric hydroxyl radicals. Vehicle emissions and heterogeneous reactions have been identified as major sources of urban HONO. Here, we report on HONO emissions from residential natural gas (RNG) for water and space heating in urban areas based on in situ measurements. The observed HONO emission factors (EFs) of RNG heating vary between 6.03 and 608 mg·m-3 NG, which are highly dependent on the thermal load. The highest HONO EFs are observed at a high thermal load via the thermal NO homogeneous reaction. The average HONO EFs of RNG water heating in winter are 1.8 times higher than that in summer due to the increased thermal load caused by the lower inlet water temperatures in winter. The power-based HONO EFs of the traditional RNG heaters are 1085 times and 1.7 times higher than those of gasoline and diesel vehicles that meet the latest emission standards, respectively. It is estimated that the HONO emissions from RNG heaters in a typical Chinese city are gradually close to emissions from on-road vehicles when temperatures decline. These findings highlight that RNG heating is a non-negligible source of urban HONO emissions in China. With the continuous acceleration of coal-to-gas projects and the continuous tightening of NOx emission standards for vehicle exhaust, HONO emissions from RNG heaters will become more prominent in urban areas. Hence, it is urgently needed to upgrade traditional RNG heaters with efficient emission reduction technologies such as frequency-converted blowers, secondary condensers, and low-NOx combustors.

Investigating the Incorporation of Idle, High Idle, and Driving Acceleration NOx Emissions Tests into the Periodic Technical Inspection Procedures

NOx pollution is one of the greatest air quality issues that urban areas face today, particularly within the European Union (EU), yet currently this pollutant is only controlled through the homologation process. There is currently no periodic technical inspection (PTI) process for NOx emissions within the EU, leaving a weakness in the legislation that is currently allowing high polluters to negatively impact air quality. Work needs to be performed to incorporate a simple, quick, inexpensive, and representative test to accurately identify these high emitters within the on-road vehicle fleet. This paper investigates options for the incorporation of a NOx test into the EU PTI test procedures. In a trial constituting over 600 vehicles, a 3DATX parSYNC was used to measure the NOx emissions over a series of short test types. These are an idle test, two types of high idle test (a constant high idle and a rapid high idle), and an on-road driving dynamic acceleration test. The repeatability of all three test types was good. The NOx concentrations have strong correlations to the mass emissions for each test type, with the use of mean concentrations being deemed more representative than the use of maximum concentrations. The mean results across the tested fleet are calculated and used to define pass/fail thresholds for different vehicle types. The findings of this work show that multiple test methods have the potential to characterize NOx emissions from a vehicle, but in order to catch high emitters on a PTI test, the unloaded idle and high idle test types are not suitable substitutes for a dynamic acceleration test, particularly for petrol vehicles.

Methodology for Mobile Toxics Deterministic Human Health Risk Assessment and Case Study

Air toxic emissions from on-road mobile sources are significant contributors to the degradation of air quality in urban and dense population centers. Research led by the United States Environmental Protection Agency (EPA) identified more than 1162 hazardous air pollutants (HAPs) in the exhaust and evaporative emissions from on-road mobile sources. However, less than 70 hazardous air pollutants are monitored by regulatory agencies. HAPs emitted from Mobile Sources are known as Mobile Source Air Toxics (MSATs). The EPA estimates that approximately half of the cancer risk and 74% of noncancer health impacts from air toxics is attributed to mobile sources. The quantification of the risk associated with MSATs exposure remains limited to date, and only a few MSATs have ambient air quality standards to protect human health and welfare. This work presents a novel and validated methodology to quantify the myriad health risks associated with exposure to on-road mobile emissions. This methodology is introduced in the form of a pipelined analysis process, which may be employed in existing and new transportation projects. The proposed new methodology integrates results from three different types of models: on-road vehicle emissions inventory models such as MOVES and IVE, air dispersion models such as AERMOD and SCIPUFF, and risk estimate models for human and ecological receptors such as the 2005 Final U.S. EPA Human Health Risk Assessment Protocol for Hazardous Waste Combustion Facilities. The result of this research work is a new methodology that provides regulators and risk analysts with a more detailed awareness of the health impacts of MSATs. A case study of Saint Paul, Minnesota, validated the air dispersion modeled results against monitored data, and the agreement was acceptable (i.e., the estimates were within a factor of two of the observations). Three high-population locations in the Saint Paul area were evaluated for human health risk, with the observation that at two of these locations, the Saint Paul—Ramsey Health Center and Anderson Office Building, the calculated cancer risk is in excess of the target risk level of 1.0E-05 for benzo(a)pyrene. The methodology presented in this paper allows regulators, risk analysts, and air quality engineers to better estimate multi-pathway cancer and noncancer risk associated with acute and chronic exposure to MSATs. Moreover, this work provides a science-based aid to policy decision makers when considering factors that most significantly affect population health and ecology.

Modelling the Impacts of Hydrogen–Methane Blend Fuels on a Stationary Power Generation Engine

To reduce greenhouse gas emissions from natural gas use, utilities are investigating the potential of adding hydrogen to their distribution grids. This will reduce the carbon dioxide emissions from grid-connected engines used for stationary power generation, and it may also impact their power output and efficiency. Promisingly, hydrogen and natural gas mixtures have shown encouraging results regarding engine power output, pollutant emissions, and thermal efficiency in well-controlled on-road vehicle applications. This work investigates the effects of adding hydrogen to the natural gas fuel for a lean-burn spark-ignited four-stroke, 8.9 liter eight-cylinder naturally aspirated engine used in a commercial stationary power generation application via an engine model developed in the GT-SUITETM modelling environment. The model was validated for fuel consumption, air flow, and exhaust temperature at two operating modes. The focus of the work was to assess the sensitivity of the engine’s power output, brake thermal efficiency, and pollutant emissions to blends of methane with 0–30% (by volume) hydrogen. Without adjusting for the change in fuel energy, the engine power output dropped by approximately 23% when methane was mixed with 30% by volume hydrogen. It was found that increasing the fueling rate to maintain a constant equivalence ratio prevented this drop in power and reduced carbon dioxide emissions by almost 4.5%. In addition, optimizing the spark timing could partially offset the increases in in-cylinder burned and unburned gas temperatures and in-cylinder pressures that resulted from the faster combustion rates when hydrogen was added to the natural gas. Understanding the effect of fuel change in existing systems can provide insight on utilizing hydrogen and natural gas mixtures as the primary fuel without the need for major changes in the engine.

The contributions of non-methane hydrocarbon emissions by different fuel type on-road vehicles based on tests in a heavily trafficked urban tunnel

Automobile exhaust is a major source of volatile organic compounds (VOCs) in metropolitan areas, yet it is difficult to accurately determine the contributions of different types of on-road vehicles. Tunnel tests are an effective way to measure real-world vehicle emissions, and the data collected are also suitable for receptor modeling to analyze the contributions of non-methane hydrocarbons (NMHCs) from different types of vehicles, as the closed environment ensures good mixing and minimal aging. In this study, tunnel tests were conducted inside a heavily trafficked city tunnel in Guangzhou in south China, and the positive matrix factorization (PMF) model was applied to the inlet-outlet incremental NMHC data. The results revealed that gasoline vehicles (GVs), Liquefied Petroleum Gas vehicles (LPGVs), and diesel vehicles (DVs) were responsible for 39 %, 45 % and 16 % of NMHCs, and 52 %, 23 %, and 24 % of the ozone formation potentials, respectively. LPGVs were the largest contributor of (56 %) alkanes, and GVs were the largest contributor of aromatics (61 %) and C2-C4 alkenes (55 %). With the video-recorded traffic counts the emissions of different fuel types are further compared on a per-vehicle-per-kilometer basis, and the results reveal that LPGVs and GVs were comparable in the OFPs of NMHCs emitted per kilometer, while on average a DV emitted 2.0 times more NMHCs than a GV with 2.4 times more OFPs. This study highlights substantial contribution of reactive alkenes and aromatics by DVs and the benefits of strengthening diesel exhaust control in terms of preventing ozone pollution.

Inferred vehicular emissions at a near-road site: Impacts of COVID-19 restrictions, traffic patterns, and ambient air temperature

Vehicles are a major source of anthropogenic emissions of carbon monoxide (CO), nitrogen oxides (NOx), and black carbon (BC). CO and NOx are known to be harmful to human health and contribute to ozone formation, while BC absorbs solar radiation that contributes to global warming and also has negative impacts on human health and visibility. Travel restrictions implemented during the COVID-19 pandemic provide researchers the opportunity to study the impact of large, on-road traffic reductions on local air quality. Traffic counts collected along Interstate-95, a major eight-lane highway in Maryland (US), reveal a 60% decrease in passenger car totals and an 8.6% (combination-unit) and 21% (single-unit) decrease in truck traffic counts in April 2020 relative to prior Aprils. The decrease in total on-road vehicles led to the near-elimination in stop-and-go traffic and a 14% increase in the mean vehicle speed during April 2020. Ambient near-road (NR) BC, CO, NOx, and carbon dioxide (CO2) measurements were used to determine vehicular emission ratios (ΔBC/ΔCO, ΔBC/ΔCO2, ΔNOx/ΔCO, ΔNOx/ΔCO2, and ΔCO/ΔCO2), with each ratio defined as the slope value of a linear regression performed on the concentrations of two pollutants within an hour. A decrease of up to a factor of two in ΔBC/ΔCO, ΔBC/ΔCO2, ΔNOx/ΔCO2, and in the fraction of on-road diesel vehicles from weekdays to weekends shows diesel vehicles to be the dominant source of BC and NOx emissions at this NR site. We estimate up to a 70% reduction in BC emissions in April 2020 compared to earlier years, and attribute much of this to lower diesel BC emissions resulting from improvements in traffic flow and fewer instances of acceleration and braking. Future efforts to reduce vehicular BC emissions should focus on improving traffic flow or turbocharger lag within diesel engines. Inferred BC emissions from the NR site also depend on ambient temperature, with an increase of 54% in ΔBC/ΔCO from −5 to 20 °C during the cold season, similar to previous studies that reported increasing BC emissions with rising temperature. The default setting of MOVES3, the current version of the mobile emission model used by the US EPA, does not adjust hot-running BC emissions for ambient temperature. Future work will focus on improving the accuracy of mobile emissions in air quality modeling by incorporating the effects of temperature and traffic flow in the system used to generate mobile emissions input for commonly used air quality models.

Vehicle Detection and Recognition Approach in Multi-Scale Traffic Monitoring System via Graph-Based Data Optimization.

Over the past few years, significant investments in smart traffic monitoring systems have been made. The most important step in machine learning is detecting and recognizing objects relative to vehicles. Due to variations in vision and different lighting conditions, the recognition and tracking of vehicles under varying extreme conditions has become one of the most challenging tasks. To deal with this, our proposed system presents an adaptive method for robustly recognizing several existing automobiles in dense traffic settings. Additionally, this research presents a broad framework for effective on-road vehicle recognition and detection. Furthermore, the proposed system focuses on challenges typically noticed in analyzing traffic scenes captured by in-vehicle cameras, such as consistent extraction of features. First, we performed frame conversion, background subtraction, and object shape optimization as preprocessing steps. Next, two important features (energy and deep optical flow) were extracted. The incorporation of energy and dense optical flow features in distance-adaptive window areas and subsequent processing over the fused features resulted in a greater capacity for discrimination. Next, a graph-mining-based approach was applied to select optimal features. Finally, the artificial neural network was adopted for detection and classification. The experimental results show significant performance in two benchmark datasets, including the LISA and KITTI 7 databases. The LISA dataset achieved a mean recognition rate of 93.75% on the LDB1 and LDB2 databases, whereas KITTI attained 82.85% accuracy on separate training of ANN.

Engine Fuel Consumption Modelling Using Prediction Error Identification and On-Road Data

Engine modelling is an important step in predicting the fuel consumption of a vehicle. Existing methods in the literature require dedicated tests on a test track or on a chassis dynamometer or they require measurements from several days of vehicle operation. This article proposes a new method to model fuel flow rate of a diesel engine and a compressed gas engine using prediction error identification and on-road data collection. The model inputs are the engine torque and speed. The on-road vehicle data was collected during normal transport operations. The identification data set was approximately 99% shorter than the baseline method. The proposed method is applicable for other types of vehicles, including electric vehicles. The identified engine models have less than 1.3% mean error and 2.5% RMS error.

A new portable open-path instrument for ambient NH3 and on-road emission measurements

Increased attentions to vehicle emission of NH3 have been paid since it is generally regarded as an important source in urban areas. Here, we developed a movable instrument based on Differential Optical Absorption Spectroscopy (DOAS) principle for detecting on-road NH3, which can avoid the losses in the sampling process attributed to the non-sampling methods. For this mobile DOAS, the temporal resolution, detection limit and relative error for NH3 were 1 min, 2.29 ppbv and 4.57% ± 2.44%, respectively. By employed to the on-road measurements along the arterial highway in Shanghai, the spatial distributions of NH3 and NO were obtained, and their dependence of traffic and road conditions were studied. The slopes of linear regression between NH3 and NO were 0.40, 0.02 and 0.07 on the Middle Ring Road, Outer Ring Road and Chongming Island Ring Road. It indicates that light gasoline vehicles (LGVs) were found to be the main contributor to NH3 emissions, while heavy-duty diesel vehicles (HDVs) mainly emitted NO. Based on the measured NH3 in the tunnel, the mileage-based NH3 emission factor per vehicle was estimated to be 17.9 ± 6.3 mg/km. The reported open-path instrument can be broadly used in on-road pollutant monitoring or vehicle emissions, and the measurements can reveal the real situation of emission characteristics, even find the abnormal operations of vehicle catalyst system.

Automotive tire defect characterization with a portable digital holographic camera with an extended field of view.

Detection of tire defects is of prime importance for the on-road safety of vehicles. Hence, a quick non-invasive mechanism is required for frequent testing of tires in service as well as for quality checks of newly manufactured tires in the automotive industry. We propose a methodology based on the principle of double-exposure digital holographic interferometry, implemented with a portable digital holographic camera for successful detection and dimension measurement of tire defects. To implement the principle, mechanical load is applied to a tire to produce interferometric fringes by comparing the normal and stressed states of the tire surface. The defects in the tire sample are identified from discontinuities in the interferometric fringes. Quantitative analysis of the displacement of the fringes gives the dimensions of the defects. Some experimental results are presented, validated by a vernier caliper.

Optical properties of vehicular brown carbon emissions: Road tunnel and chassis dynamometer tests

Brown carbon (BrC), as an important light-absorbing aerosol, significantly impacts regional and global climate. Vehicle emission is a nonnegligible source of BrC, but the optical properties of BrC emitted from vehicles remain poorly understood. This study evaluates the absorption Ångström exponent (AAE) of traffic-related light-absorbing aerosols (i.e., AAETr) and the absorption emission factor (EFabs) of vehicular BrC via chassis dynamometer tests and a road tunnel measurement in Tianjin, China. AAETr are estimated as 0.98–1.33 and 1.11 ± 0.001 for tested vehicles and on-road vehicle fleet, respectively. The AAE of vehicular BrC (AAEBrC) is 3.83 ± 0.092 for on-road vehicle fleet. The vehicle technology updates effectively reduce the EFabs of vehicular BrC. Among the four tested China 5 and China 6 gasoline vehicles in the chassis dynamometer tests, BrC EFabs of China 5 gasoline direct injection vehicle is the highest, while China 6 mixing fuel injection vehicle exhibits the lowest EFabs. The BrC EFabs of on-road vehicle fleet at 370 nm wavelength are 0.081 ± 0.0058 m2 kg−1 for mixed fleet, 0.074 ± 0.018 m2 kg−1 for gasoline vehicles (GVs), and 1.66 ± 0.71 m2 kg−1 for diesel vehicles (DVs) in the tunnel measurement. EFabs of GV fleet in the road tunnel is higher than China 5 and China 6 vehicles, as China 1–4 vehicles accounted for 26.8% of the total vehicle fleet in the tunnel. EFabs of vehicular BrC are lower than those from biomass burning and coal combustion emissions. The light absorption of BrC from GVs and DVs accounts for 7.2 ± 2.1% and 1.5 ± 0.77% of total traffic-related absorption at 370 nm, respectively. Our study provides optical features of BrC from vehicle source and could contribute to estimating the impacts of vehicular aerosol emissions on global and regional climate change.

TAKEN: A Traffic Knowledge-Based Navigation System for Connected and Autonomous Vehicles

Connected and autonomous vehicles (CAVs) have witnessed significant attention from industries, and academia for research and developments towards the on-road realisation of the technology. State-of-the-art CAVs utilise existing navigation systems for mobility and travel path planning. However, reliable connectivity to navigation systems is not guaranteed, particularly in urban road traffic environments with high-rise buildings, nearby roads and multi-level flyovers. In this connection, this paper presents TAKEN-Traffic Knowledge-based Navigation for enabling CAVs in urban road traffic environments. A traffic analysis model is proposed for mining the sensor-oriented traffic data to generate a precise navigation path for the vehicle. A knowledge-sharing method is developed for collecting and generating new traffic knowledge from on-road vehicles. CAVs navigation is executed using the information enabled by traffic knowledge and analysis. The experimental performance evaluation results attest to the benefits of TAKEN in the precise navigation of CAVs in urban traffic environments.

Closed-loop Control of SI←→RCCI Mode Transitions in a Multi-Mode Combustion Engine

The primary barrier to implementing the RCCI mode in on-road vehicles is its limited operating range due to excessive pressure rise rates and complexity of combustion control during mode switching. The feasible operating range of the RCCI mode is only a subset of the speed-load range of conventional spark ignition (SI) engines. Therefore, a multi-mode engine concept operating in RCCI and SI modes can cover required engine operating range and can improve engine performance in terms of thermal efficiency and emissions. The goal of this study is to develop a model-based closed loop control of an SI-RCCI multi-mode engine. To this end, a gain scheduled model predictive controller (MPC) is developed for SI-RCCI-SI mode switching. A Kalman filter is designed for state estimations. The control architecture includes; a supervisory controller that determines the optimal operating mode based on the requested engine load and speed, and gain scheduled MPC and a Kalman filter. The controller performance is validated on a 2-liter multi-mode engine for SI-RCCI-SI mode switching at different engine loads (NMEP) and combustion phasings (CA50). The controller is capable of tracking NMEP and CA50 in both modes and during mode transitions while maintaining air-fuel ratio near stoichiometry in SI mode and constraining the maximum pressure rise rate (MPRR) below 8 bar/CAD in the RCCI mode.

Estimating Battery Size Requirements for Tractor Electrification of Row-Crop Planting Operations

Highlights Calculate and compare engine power needed for row-crop planting from ASAE standards and reported CAN data. Estimate energy requirements and the equivalent electric battery’s volume and mass for row-crop tractor. Discuss the feasibility for electrification of a fully electric row-crop tractor. Abstract. Power sources such as batteries, used for both on-road and off-road vehicles, are advancing at a rapid pace. Electric batteries are becoming more power dense, thus allowing them to be used as a power source to replace previous diesel or gasoline-powered systems. Efforts are underway to transition off-road agricultural vehicles from Internal Combustion Engine (ICE) vehicles to electric vehicles (EVs); however, the energy requirements of typical agricultural field operations need to be fully understood before such a transition can occur. Theoretical prediction equations available in the American Society of Agricultural and Biological Engineers (ASABE) standards or the use of engine load data from a tractor’s Controller Area Network (CAN) bus are two methods for determining the energy demands of implements on tractors. In this study, tractor CAN bus data was collected from multiple no-till corn planting operations to estimate the energy requirements of the planting operation in kWh. The estimated energy requirement was used to determine the equivalent physical mass and volume of a lithium-ion battery needed to power a hypothetical fully electric tractor for comparable planting operations. The estimated battery capacities using the worst-case field-use scenario in this study were 1117 kWh for operating a 16-row planter to plant 68 ha (168 acre) in a day at an average ground speed of 8.6 km h-1 (5.4 mph) for 14 hours, and 2658 kWh for operating a 48-row planter to plant 158 ha (391 acre) in a day at an average ground speed of 8.9 km h-1 (5.6 mph) for 15 hours. Given the current battery energy density requirements, the equivalent battery masses and volumes were found to be 5,319 kg and 3.33 m3, 12,657 kg and 7.93 m3, for 16-row planter and 48-row planter, respectively. These high kWh estimates needed to power future fully-electric tractor power units are based on worst-case scenarios that could be encountered in real field operations that use wide planters over long operating hours. Keywords: Battery, Controller Area Network, Electric Vehicles, Electrification, No-till, Planting, Row-crop, Tractor.

Mitigation of emissions and energy consumption due to light-duty vehicle size increases

Analysts need to consider the shift of light-duty vehicle sales from cars to light trucks to accurately project fuel consumption, greenhouse gas emissions, and consumer spending. We find that energy consumed by on-road light-duty vehicles in the United States can vary by 10% by changing assumptions regarding the sales mix. Scenarios aligned with third-party forecasts, assuming a greater sales proportion of light trucks and fewer cars, yield petroleum consumption 3–8% higher than forecasts from the U.S. Energy Information Administration (EIA), with greenhouse gas emissions 5–7% higher and a 4–9% increase in consumer spending on vehicles and fuel. This incremental energy consumption can be offset by additional technologies for these vehicles. If most sedans were phased out in favor of larger sport utility vehicles, we find a sales share of 30% battery electric vehicles or 39% hybrid electric vehicles would equal the emissions of the EIA Reference Case.

Neural-Network-Optimized Vehicle Classification Using Clustered Image and Fiber-Sensor Datasets

Internet of Things (IoT) becomes indispensable for transport and automotive industry to advance functions in on-road traffic monitoring. Indeed, smart management tools and machine learning concepts are inevitable in vehicle categorization systems. However, existing systems are only built on individual platforms, and for a most part, the classification accuracy remains limited. In this work, these challenges are tackled by designing a novel convolutional neural network (CNN) that substantially improves on-road vehicle classification. In particular, we experimentally harness, to the best of our knowledge for the first time, two different datasets from separated technological platforms based on close-circuit television (CCTV) and fiber Bragg grating (FBG) sensors, respectively. Standard neural networks in single FBG platform yield limited classification accuracy of only 34% - 62% for <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">AlexNet</i> , 51% - 77% for <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GoogleNet</i> , 57% - 78% for <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ResNet-50</i> , and 59% - 86% for <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ResNet-101</i> . In contrast, hybrid CNN classification with individual CCTV and FBG datasets substantially improves detection levels, reaching in-class accuracy up to 90% - 97%. Moreover, this classification concept includes an intrinsic back-up verification with respect to each platform compensating the shortcomings of individual technologies. Our demonstration can make key advances towards near-unity accuracy in vehicle classifications for IoT systems, capitalizing on cost-effective and well-established platforms.

Characteristics of fuel cells under static and dynamic conditions

Modern internal combustion powertrains are the main source of propulsion for on-road and non-road vehicles. However, they are increasingly being replaced by electric or fuel cell-equipped alternative propulsion systems. The article presents a study of fuel cell characteristics operating under both static and dynamic conditions, with a 1.2 kW fuel cell set with a voltage converter and lead-acid batteries. In the conducted tests, the fuel cell stack's maximum efficiency reached 65%. Load tests (static and dynamic) have indicated higher fuel cell efficiencies when using hybrid operation with a DC/DC converter and battery.

Combined effects of vehicles and waste incineration on urban air halogenated and parent polycyclic aromatic hydrocarbons

Traffic emissions and waste incineration are the main sources of PAHs in urban atmosphere, but their spatially superimposed effects are currently unclear. This study assessed the spatial distribution of PAHs and HPAHs concentrations in the atmosphere of Shenzhen by simulating the spatial and temporal dispersion of PAHs and HPAHs emissions from on-road vehicles and municipal solid waste incinerators (MSWIs). Generally, the concentrations of PAHs and HPAHs were higher on workdays than on weekends due to higher traffic volumes, while the prevailing wind direction of the northeast could cause more widespread dispersion of PAHs and HPAHs within Shenzhen’s atmosphere. After superimposing the spatial distribution of pollutants emitted by vehicles and MSWIs, PAHs within 1000 m downwind of MSWIs are mainly contributed by MSWIs and beyond 1000 m by vehicles. The cancer risk values induced by exposure to PAHs and HPAHs via inhalation in Shenzhen were below the acceptable risk level for males and females in each age group, while adults faced the highest cancer risk, followed by adolescents and children. However, spatially, the cancer risk values were above the priority risk level for adult males in localized high-traffic areas in Futian and Luohu districts.