Driving Pattern Recognition Research Articles

Variable horizon multivariate driving pattern recognition framework based on vehicle-road two-dimensional information for electric vehicle

Driving condition is the core factor influencing vehicle energy management and power output strategies, hence realistic and accurate driving pattern recognition is crucial for realizing intelligent control of Electric Vehicles in complex scenarios. However, driving patterns mismatch with real scenarios caused by lack of road environment information and the reduced global recognition accuracy triggered by the fixed recognition horizon together constrain the development of intelligent control strategies. In this study, a framework for variable horizon multivariate driving pattern recognition strategy based on vehicle-road two-dimensional (2D) information is proposed. Specifically, multivariate driving pattern extraction is carried out based on real vehicle-road 2D data, and the extraction data serves as input for training offline driving pattern recognizer. To meet the challenge of large differences and frequent changes in driving conditions during the online application, a variable horizon recognition strategy is proposed based on the optimal horizon matching regularity. Finally, test results show that the proposed approach achieves an average recognition accuracy of 88.42% under different cycles, including 88.39% for complex driving cycles. Furthermore, the recognition accuracy is improved by 63.14% and 47.51% compared with the fixed long/short horizon. The framework can provide a basis for intelligent control strategies, thus broadening the Electric Vehicle application scenarios.

Optimal Energy Management Strategy Based on Driving Pattern Recognition for a Dual-Motor Dual-Source Electric Vehicle

Optimal Energy Management Strategy Based on Driving Pattern Recognition for a Dual-Motor Dual-Source Electric Vehicle

Adaptive power allocation strategy for hybrid energy storage system based on driving pattern recognition

The power allocation strategy of hybrid energy storage systems plays a decisive role in energy management for electric vehicles. However, existing online real-time power allocation strategies primarily rely on expert knowledge to make rules. Due to the real time changes in driving patterns, it is necessary for the power allocation strategy to possess adaptive capabilities to ensure optimal energy efficiency. Therefore, an adaptive power allocation strategy is proposed in this paper, which takes into account the driving patterns of electric vehicles. Firstly, driving pattern recognition is performed using the K-means clustering method to extract features of driving profiles. Then, the fuzzy logic control strategy is adopted to make the rules of filter frequency, and the rules is adaptive adjusted based on the results of driving pattern recognition. To achieve smoother power output from lithium-ion batteries and prolong their lifespan, the digital low pass filter is used to accomplish the fuzzy logic control strategy, and the filtering frequency is online optimized by the above fuzzy controller. According to the digital low pass filter, the load power demand is decomposed high-frequency component and low-frequency component. The decomposed high-frequency component of power demand is supplied by supercapacitors, while the low-frequency component is provided by lithium-ion batteries. To validate the effectiveness of the proposed method, comparative experiments are conducted between the proposed strategy and another four strategies. The experimental results demonstrate that the proposed strategy can effectively smooth the power output of lithium-ion batteries and reduce total energy losses.

Adaptive Fuzzy Power Management Strategy for Extended-Range Electric Logistics Vehicles Based on Driving Pattern Recognition

The primary objective of an energy management strategy is to achieve optimal fuel economy through proper energy distribution. The adoption of a fuzzy energy management strategy is hindered due to different reasons, such as uncertainties surrounding its adaptability and sustainability compared to conventional energy control methods. To address this issue, a fuzzy energy management strategy based on long short-term memory neural network driving pattern recognition is proposed. The time-frequency characteristics of vehicle speed are obtained using the Hilbert–Huang transform method. The multi-dimensional features are composed of the time-frequency features of vehicle speed and the time-domain signals of the accelerator pedal and brake pedal. A novel driving pattern recognition approach is designed using a long short-term memory neural network. A dual-input and single-output fuzzy controller is proposed, which takes the required power of the vehicle and the state of charge of the battery as the input, and the comprehensive power of the range extender as the output. The parameters of the fuzzy controller are selected according to the category of driving pattern. The results show that the fuel consumption of the method proposed in this paper is 5.8% lower than that of the traditional fuzzy strategy, and 4.2% lower than the fuzzy strategy of the two-dimensional feature recognition model. In general, the proposed EMS can effectively improve the fuel consumption of extended-range electric vehicles.

Analysing the Energy Consumption of Urban Electric Vehicles for Real Driving Scenarios

Driving efficiently with an electric vehicle is a key requirement to reduce energy consumption. This article aims to develop a unique two-layer driving pattern recognition approach for the classification of roadway type along with traffic situation for a particular route. The main idea is to use this information to accurately predict the driving pattern and hence analyse the energy consumption of an electric vehicle driving along a specific route under real driving scenarios. The main novelty of the contribution to develop a classification technique based on representative drive cycles, which characterise a roadway type and traffic condition on the basis of certain features. The simulation results indicate a classification accuracy of 85.2–97.7% in driving pattern recognition resulting in prediction of energy consumption under real traffic and road conditions.

Review of Energy Management Methods for Fuel Cell Vehicles: From the Perspective of Driving Cycle Information.

Energy management methods (EMMs) utilizing sensing, communication, and networking technologies appear to be one of the most promising directions for energy saving and environmental protection of fuel cell vehicles (FCVs). In real-world driving situations, EMMs based on driving cycle information are critical for FCVs and have been extensively studied. The collection and processing of driving cycle information is a fundamental and critical work that cannot be separated from sensors, global positioning system (GPS), vehicle-to-vehicle (V2V), vehicle-to-everything (V2X), intelligent transportation system (ITS) and some processing algorithms. However, no reviews have comprehensively summarized the EMMs for FCVs from the perspective of driving cycle information. Motivated by the literature gap, this paper provides a state-of-the-art understanding of EMMs for FCVs from the perspective of driving cycle information, including a detailed description for driving cycle information analysis, and a comprehensive summary of the latest EMMs for FCVs, with a focus on EMMs based on driving pattern recognition (DPR) and driving characteristic prediction (DCP). Based on the above analysis, an in-depth presentation of the highlights and prospects is provided for the realization of high-performance EMMs for FCVs in real-world driving situations. This paper aims at helping the relevant researchers develop suitable and efficient EMMs for FCVs using driving cycle information.

Meta rule-based energy management strategy for battery/supercapacitor hybrid electric vehicles

Driving pattern recognition (DPR) is widely used to improve the robustness of rule-based energy management strategies (EMS). However, the number and quality of preset patterns limit further performance improvements. This paper proposes a meta rule-based EMS to replace the role of DPR, wherein the meta rule refers to a rule that determines the parameters of the energy management rule. Firstly, the power distribution rule pattern is derived through the analysis of dynamic programming results. Then, the value of the parameters is obtained through clustering. Subsequently, a feature screening method based on mutual information is proposed to select useful features of driving conditions and system states. The seagull optimization algorithm is then employed to find the mathematical expression of the meta rule. Finally, simulation results demonstrate the effectiveness of the meta rule-based EMS. Compared with DPR-based and long short-term memory-based EMS, the meta rule-based EMS can reduce the Ampere-hour throughput of Li-ion batteries by 17.0% and 9.7% respectively under the China light-duty vehicle test cycle, and by 3.4% and 16.6% respectively under a real-world driving condition. The meta rule-based EMS requires only dozens of milliseconds for each moment and is suitable for real-time systems.

An Improved Data-Driven Method for Steering Feedback Torque of Driving Simulator

Traditional modeling methods of steering feedback torque (SFT) widely used in driving simulator are mainly physics-based, but suffers from high model complexity and low model accuracy due to the inaccurate or unknown structure, dynamics, and related parameters of the steering system, such as backlash or dry friction. For these problems, data-driven methods were considered to have advantageous in modeling such a complex yet dynamic system. However, due to the limitation of the data collected for model training from real vehicle test, the commonly used artificial neural network (ANN) cannot fully reflect the steering dynamics, and does not work well in many cases where collected data do not cover, and thus lead to poor generalization performance. In this article, a long short-term memory (LSTM) model is proposed and proved to simulate well on the dynamics of SFT, and works well in most driving conditions. A driving simulator is developed to verify the performance of different models in terms of generalization of working conditions. Based on the driving simulator data, we found that ANN failure generally arises in the transition between conventional and unconventional operating conditions, which is recognized by the proposed driving pattern recognition method. The LSTM-based modeling method can effectively solve this problem. The comparative analysis reveals that the improved LSTM model can simulate the SFT with higher accuracy and broader adaptability to driving simulator conditions.

Time-series clustering for pattern recognition of speed and heart rate while driving: A magnifying lens on the seconds around harsh events

Driving pattern recognition has been applied for the purposes of driving styles identification and harsh driving events detection. However, the evolution of driving behavior around and especially before such events has not been investigated at a microscopic level. The objective of this research is to reveal existing driving patterns around harsh events at the driving ‘pulse’ level i.e. a few seconds before and after the event. For that purpose, a time-series clustering approach is applied on speed and heart rate metrics of individual drivers using data collected from a large naturalistic driving study. Results show that there are distinct speed patterns before harsh braking, harsh acceleration, and harsh cornering events. A deceleration is identified shortly before most harsh acceleration and cornering events, which possibly indicates reckless behavior, i.e. drivers not dedicating enough time to smoothly brake before cornering, or of a brief ‘decision-making’ moment before the harsh manoeuvre. On the contrary, speed seems to be steady before harsh braking events. Regarding heart rate, the analysis revealed certain patterns only after raw data were cleansed and filtered. These patterns may show increasing, decreasing or variable heart rate trends, which may correspond to different stress patterns of drivers around harsh events. Finally, we introduce the concept of driving pattern consistency, which can reveal the share of individual drivers that follow the same harsh event pattern. It is indicated that more than half of the drivers are not consistent, suggesting that driving patterns around harsh events may be more context-related than driver personality-related.

Predictive hierarchical eco-driving control involving speed planning and energy management for connected plug-in hybrid electric vehicles

The connected vehicle technique has offered great opportunities to improve further plug-in hybrid electric vehicles (PHEVs) fuel economy. In this context, a predictive hierarchical eco-driving control scheme is proposed for connected PHEVs under a car-following scenario containing a cloud-layer speed planner and vehicle-layer energy management. In the cloud layer, based on the traffic data separately processed by the dynamic programming (DP) algorithm and k-means clustering method, the reference state of charge (SOC) model, the energy consumption model, the equivalent factor model and the variable-horizon speed predictor can be constructed, respectively. And for ensuring safety, comfort and fuel economy in the car-following process, the energy and SOC models are used separately as the index and state equation in the data-driven speed planning problem. Then, with the driving pattern recognition technique, the power allocation under the planned speed can be conducted by integrating the adaptive equivalent consumption minimization strategy (ECMS) with the model predictive control (MPC), thus guaranteeing fuel economy, adaptability and near-global optimality with high computational efficiency. Compared with other strategies, the effectiveness and advantages of the proposed scheme are validated in the joint simulation platform of MATLAB/Simulink and GT-SUITE.

Online adaptive energy management strategy for fuel cell hybrid vehicles based on improved cluster and regression learner

An effective energy management strategy (EMS) is essential to ensure the safe and efficient operation of the fuel cell hybrid vehicles. In this paper, an online adaptive EMS is proposed for the fuel cell hybrid vehicles to minimize hydrogen consumption and adjust the strategy according to the driving conditions. Driving pattern recognition is realized by the improved k-means cluster approach which combines multiple k-means clusters with specific distances to serve as the classifier. In each driving pattern, separate machine learning (ML) models are trained to obtain the energy management regression learner. Comparison experiments are performed to determine the optimal ML model and input parameters. The effectiveness of the proposed EMS is evaluated using two compound test driving cycles. Results show that the proposed method achieves the lowest fuel consumption compared to the other five algorithms considered. Remarkably, it reduces hydrogen consumption by up to 5.66% when compared to commonly used methods.

A dynamic programming-optimized two-layer adaptive energy management strategy for electric vehicles considering driving pattern recognition

To achieve optimal real-time power allocation in electric vehicles, a two-layer adaptive dynamic programming (DP) optimization energy management strategy (EMS) has been proposed. The upper layer uses learning vector quantization (LVQ) to produce real-time driving pattern recognition (DPR) results. The method determines 10 characteristic parameters for training the recognition network and the length of the sampling window is 120 s. The typical driving cycles are divided into different levels of blocks to identify the real-time DPR level. The lower layer adopts the optimization strategy of DP extraction to adjust the power distribution between the battery pack and the supercapacitor pack according to the recognition results. DP is used to define a cost function to minimize the energy loss of the hybrid energy storage system (HESS) and optimize the battery usage range in the system. The near-optimal real-time EMS is extracted by analyzing the DP control behavior of the battery under the layered state of charge (SOC). The simulation results indicate that the proposed new rule control based on DP optimization (NRB) EMS improves the system efficiency by 10 % compared with the original rule-based (RB) EMS under different temperatures and DPR levels. In addition, the system efficiency gap is controlled at 3 % compared with DP.

Model Predictive Control Based Energy Management Strategy of Series Hybrid Electric Vehicles Considering Driving Pattern Recognition

This paper proposes an energy management strategy for a series hybrid electric vehicle based on driving pattern recognition, driving condition prediction, and model predictive control to improve the fuel consumption while maintain the state of charge of the battery. To further improve the computational efficiency, the discretization and linearization of the model is conducted, and the MPC problem is transferred into a quadratic programming problem, which can be solved by the interior point method effectively. The simulation is carried out by using Matlab/Simulink platform, and the simulation results verify the feasibility of the condition prediction method and the performance of the proposed method. In addition, the predictive control strategy successfully improves the fuel economy of the hybrid vehicle compared with the rule-based method.

Driver Profile and Driving Pattern Recognition for Road Safety Assessment: Main Challenges and Future Directions

This study reviews the Artificial Intelligence and Machine Learning approaches developed thus far for driver profile and driving pattern recognition, representing a set of macroscopic and microscopic behaviors respectively, to enhance the understanding of human factors in road safety, and therefore reduce the number of crashes. It provides a definition of the two scientific fields in terms of safety, and identifies the most efficient approaches used regarding methodology, data collection and driving metrics. Results show that K-means and Neural Networks are the most commonly used methodologies for driver profile identification, and Dynamic Time Warping for driving pattern detection. Most studies discovered driver profiles related to aggressiveness, considering mainly speed and acceleration as driving metrics. Based on the gaps and challenges identified, this paper provides a new framework for combining microscopic and macroscopic driving behavior analysis, instead of examining them separately as is the state-of-theart. Such combined results can potentially improve the development of traffic risk models, which could be exploited in applications that monitor drivers in real-time and provide feedback. These models will represent human behavior more accurately, which can eventually lead to the recognition of “optimal” human driving patterns that Automated Vehicles (AV) could ‘mimic’ to become safer.

Discriminative Transfer Learning for Driving Pattern Recognition in Unlabeled Scenes

Abstract: Driving behavior has a large impact on behaviour of driver. The lack of a labeled data problem in a driving scene substantially hinders the driving pattern recognition accuracy. However, modeling driving behavior under the dynamic driving conditions is complex, making a quantitative analysis of the the driving behaviour. In this paper, the Driver behaivor dataset was collected from dataset repository. Then, we have to implement the pre-processing techniques. Then, the system is developed the machine learning algorithm such as Random forest and Support Vector Machine algorithm. The experimental results shows that some performance metrics such as accuracy, precision, recall and f1-score. By using the ML is, to classify the Aggressive, normal, slow.

Encoder–Decoder-Based Velocity Prediction Modelling for Passenger Vehicles Coupled with Driving Pattern Recognition

To improve the performance of predictive energy management strategies for hybrid passenger vehicles, this paper proposes an Encoder–Decoder (ED)-based velocity prediction modelling system coupled with driving pattern recognition. Firstly, the driving pattern recognition (DPR) model is established by a K-means clustering algorithm and validated on test data; the driving patterns can be identified as urban, suburban, and highway. Then, by introducing the encoder–decoder structure, a DPR-ED model is designed, which enables the simultaneous input of multiple temporal features to further improve the prediction accuracy and stability. The results show that the root mean square error (RMSE) of the DPR-ED model on the validation set is 1.028 m/s for the long-time sequence prediction, which is 6.6% better than that of the multilayer perceptron (MLP) model. When the two models are applied to the test dataset, the proportion with a low error of 0.1~0.3 m/s is improved by 4% and the large-error proportion is filtered by the DPR-ED model. The DPR-ED model performs 5.2% better than the MLP model with respect to the average prediction accuracy. Meanwhile, the variance is decreased by 15.6%. This novel framework enables the processing of long-time sequences with multiple input dimensions, which improves the prediction accuracy under complicated driving patterns and enhances the generalization-related performance and robustness of the model.

Intelligent Feature Selection Using GA and Neural Network Optimization for Real-Time Driving Pattern Recognition

Driving cycles have a great influence on vehicles’ fuel economy, control performance and drivability. In this paper, vehicle speed is considered and twelve statistical features for driving pattern recognition are selected to obtain the characteristics of driving cycles. To extract the statistical features online from the speed distribution information, the sampling and updating windows are set and the number of features is reduced. Moreover, since the structure and parameters of neural network are crucial to the classifier, the structure and parameters of neural network, the sampling and updating window size, the feature subset selection are simultaneously optimized by genetic algorithm (GA) to improve the classifying accuracy and simplify neural network structure. The hybrid encoding/decoding and the structure operator are designed to optimize the whole neural network classifier. Four typical driving patterns, i.e., congested urban road, flowing urban road, suburban and highway, are selected based on multiple driving cycles. Simulation results show that the classifiers with GA optimized features have more powerful classification capability than principle component analysis (PCA) and kernel PCA (KPCA) based on k-nearest neighbor, support vector machine neural network classifiers and KPCA Convolutional neural network classifier. The proposed classifier obtains the satisfying classification accuracy with faster real-time classifying speed.

Numerical Assessment of Auto-Adaptive Energy Management Strategies Based on SOC Feedback, Driving Pattern Recognition and Prediction Techniques

The Equivalent Consumption Minimization Strategy (ECMS) is a well-known control strategy for the definition of optimal power-split in hybrid-electric vehicles, because of its effectiveness and reduced calibration effort. In this kind of Energy Management Systems (EMS), the correct identification of an equivalence factor (K), which translates electric power in equivalent fuel consumption, is of paramount importance. To guarantee charge sustaining operation, the K factor must be adjusted to different mission profiles. Adaptive ECMS (A-ECMS) techniques have thus been introduced, which automatically determine the optimal equivalence factor based on the vehicle mission. The aim of this research activity is to assess the potential in terms of fuel consumption and charge sustainability of different A-ECMS techniques on a gasoline hybrid-electric passenger car. First, the 0D vehicle and powertrain model was developed in the commercial CAE software GT-SUITE. An ECMS-based EMS was used to control the baseline powertrain and three alternative versions of an auto-adaptive algorithm were implemented on top of that. The first A-ECMS under study was based on feedback from the battery State of Charge, while the second and third on a Driving Pattern Recognition/Prediction algorithm. Fuel consumption was assessed using the New European Driving Cycle (NEDC), the Worldwide Harmonized Light Vehicles Test Cycle (WLTC) and Real Driving Emissions (RDE) driving cycles by means of numerical simulation. A potential improvement of up to 4% Fuel Economy was ultimately achieved on an RDE driving cycle with respect to the baseline ECMS.

Multi-objective energy management strategy for fuel cell hybrid electric vehicle based on stochastic model predictive control

The fuel cell hybrid electric vehicle (FCHEV) has earned great interest among the automotive industry in recent years. However, the power allocation strategy between proton exchange membrane fuel cell (PEMFC) and lithium-ion battery remains a technological challenge. To conquer this problem, a multi-objective predictive energy management strategy (EMS) based on model predictive control (MPC) is proposed in this paper, combined with velocity forecast and driving pattern recognition. The comparative study is conducted to reveal the interaction between each optimization objectives. Simulation results illustrate that the proposed EMS could maintain SOC around reference, reduce fuel consumption by 6.67%, and avoid PEMFC degradation which caused by frequent start-off and rapid load change.

Integrating intelligent driving pattern recognition with adaptive energy management strategy for extender range electric logistics vehicle

Improving the fuel economy and optimizing power allocation of the extender range electric logistics vehicles importantly relies on reasonable energy management strategies and accurate driving pattern recognition. Herein, an adaptive equivalent consumption minimization strategy for extender range electric logistics vehicles is proposed. Firstly, a K-means clustering algorithm has been utilized to classify driving blocks. A novel driving pattern recognition approach is designed by using combination of variational mode decomposition (VMD) and extreme learning machine (ELM). Subsequently, a state-of-charge reference planning approach is proposed based on traffic information with dynamic programing for guiding battery energy allocation. Combining the driving pattern recognition results with the reference state-of-charge value, the proposed method acquires the optimal control action through minimizing the objective function. A hardware-in-the-loop test platform by Model Based Design (MBD) approach is implemented to validate the performance of controller. It has been verified that the proposed method can conserve the equivalent fuel consumption by respectively over 6.04% and 3.79%, and suppress the average battery power transients by over 11.61% and 2.85%, compared with the conventional energy management strategies.