Steering Speed Research Articles

MFE-SSNet: Multi-Modal Fusion-Based End-to-End Steering Angle and Vehicle Speed Prediction Network

AbstractIn the field of autonomous vehicles, accurately predicting steering angle and speed is a pivotal task. This task affects the accuracy of the final decision of the autonomous vehicle and is the basis for ensuring the safe and efficient operation of the autonomous vehicle. Previous studies have often relied on data from only one or two modalities to make predictions for steering angle and vehicle speed, which were often inadequate. In this paper, the authors propose a Multi-Modal Fusion-Based End-to-End Steering Angle and Vehicle Speed Prediction Network (MFE-SSNet). The network innovatively extends the one-stream and two-stream structure to a three-stream structure and cleverly extracts features of images, steering angles, and vehicle speeds using HRNet and LSTM layers. In addition, in order to fully fuse the feature information of different modal data, this paper also proposes a local attention-based feature fusion module. This module improves the fusion of different modal feature vectors by capturing the interdependencies in the local channels. Experimental results demonstrate that MFE-SSNet outperforms the current state-of-the-art model on the publicly available Udacity dataset.

Game Theory-Based Interactive Control for Human–Machine Cooperative Driving

To address the safety and efficient driving issues of human–machine shared control vehicles (HSCVs) in future complex traffic environments, this paper proposes a game theory-based interactive control method between HSCVs and surrounding autonomous vehicles (SVs) and involves considering different driving behaviors. In HSCV, a comprehensive driver model integrating steering control and speed control is designed based on the brain emotional learning circuit model (BELCM), and the control authority between the driver and the automation system is dynamically allocated through the evaluation of the driving safety field. Factors such as driving safety and travel efficiency that reflect personalized driving style are considered for modeling the uncertain behavior of SVs. In the interaction between HSCVs and SVs, a method based on game theory and distributed model predictive control (DMPC) that considers the uncertainty of SVs’ driving behavior is established and is finally integrated into a multi-objective constraint problem. The driver control input in an HSCV will also be introduced into the solution process. To demonstrate the feasibility of the proposed method, two test scenarios considering the driving characteristics of different SVs are established. The test results show that automation control systems can promptly stop the human driver’s dangerous operations in an HSCV and safely interact with multiple AVs with different driving characteristics.

Bearing-Based Formation Tracking Control of Nonholonomic Mobile Agents With a Persistently Exciting Leader

This paper investigates the leader-following formation control problem of nonholonomic agents with a persistently exciting leader based on bearing measurements. A follower needs to maintain a desired relative position to the leader in its local coordinate frame using bearing measurements. An integrated method consisting of a bearing-based estimator and a tracking controller is proposed to achieve the leader-following control objective. The estimator is designed to estimate the relative position to the leader in the follower's local coordinate frame. Based on the estimated relative position, the controller is proposed to drive the follower to achieve the desired formation. Different from existing integrated methods, our method does not require any conditions on the relative bearing between the leader and the follower in the global frame, which are difficult to guarantee. We characterize the conditions locally on the leader only, namely, conditions on the leader's linear and angular velocities with persistently exciting steering speed, such that the overall system is asymptotically stable with our integrated method. Besides, the proposed method is extended to the multi-agent formation control problem. A reasonable solution is proposed to the multi-agent case with a minimal leader follower structure. Simulations prove the effectiveness of the proposed methods in both cases.

Simulation and Validation of an 8 × 8 Scaled Electric Combat Vehicle

In this research, an 8 × 8 scaled electric combat vehicle (SECV) is built. The scaled vehicle is evaluated in both experimental and simulated methods to analyze its performance. The scaled vehicle is developed to apply the Ackermann condition by implementing the individual steering and individual wheel speed control system at low speed. Individual eight-wheel rotational velocity control and individual eight-wheel steering angle control in real time are developed and installed on the remotely controlled scaled vehicle to meet a perfect Ackermann condition. Three different steering scenarios are developed and applied: a traditional steering scenario (first and second axle steering), fixed third axle steering scenario (first, second, and fourth axle steering), and all-wheel steering scenario. Stationary evaluation, turn radius evaluation, and double lane change evaluation are conducted to verify the application of the Ackermann condition. The differences between the experimental results and the simulated data are within an acceptable range. An important demonstration of this research is the novel validation of physical and simulated data in the application of the Ackermann condition for eight-wheel steering and velocity control for the three steering scenarios.

Comparing real and simulated performance for an off‐road autonomous ground vehicle in obstacle avoidance

AbstractThis field report presents the results of a study of obstacle detection and avoidance (ODOA) by an autonomous ground vehicle (AGV) in off‐road driving conditions. This study included both real and simulated testing of the AGV and served as the third and final phase of a 3‐year research project studying the influence of environmental conditions over autonomous driving. We compare and contrast the results of the real and experimental field testing and report our findings on the influence of soft soil in ODOA performance by an AGV. We find that rutting in soft soil results in higher throttle effort but lower steering effort and speed. The results of these experiments will inform future studies of autonomous driving on off‐road environments.

Designing a new integrated control solution for electric power steering systems based on a combination of nonlinear techniques.

An EPS system is used to improve the stability and safety of the car when steering while also simplifying the steering process. This article introduces a novel control solution for the EPS system called BSSMCPID. This algorithm combines two nonlinear techniques, BS and SMC, with the input signal corrected by a PID technique. This algorithm provides three new contributions compared to existing algorithms: reducing system errors and eliminating phase differences, ensuring stability even when exposed to external disturbances, and reducing power consumption. The system's stability is evaluated according to the Lyapunov criterion, while the algorithm's performance is evaluated based on numerical simulation results. According to the article findings, the RMS error of the steering column angle and steering motor angle values (controlled objects) is approximately zero, and the RMS error of the steering column speed and steering motor speed is about 0.01 rad/s, which is much lower than the results obtained with traditional BS and PID controllers. When the EPS system is controlled by the integrated nonlinear method proposed in this work, the output values always closely follow the reference values with negligible errors under all investigated conditions. Additionally, power steering performance increases as speed decreases or driver torque increases, which follows the ideal assisted power steering curve. In general, the responsiveness and stability of the system are always ensured when applying this algorithm.

Beam delivery characteristics of the Hitachi carbon ion scanning system at Osaka Heavy Ion Medical Accelerator in Kansai (HIMAK).

Using the pencil beam raster scanning method employed at most carbon beam treatment facilities, spots can be moved without interrupting the beam, allowing for the delivery of a dose between spots (move dose). This technique is also known as Dose-Driven-Continuous-Scanning (DDCS). To minimize its impact on HIMAK patient dosimetry, there's an upper limit to the move dose. Spots within a layer are grouped into sets, or "break points," allowing continuous irradiation. The beam is turned off when transitioning between sets or at the end of a treatment layer or spill. The control system beam-off is accomplished by turning off the RF Knockout (RFKO) extraction and after a brief delay the High Speed Steering Magnet (HSST) redirects the beam transport away from isocenter to a beam dump. The influence of the move dose and beam on/off control on the dose distribution and irradiation time was evaluated by measurements never before reported and modelled for Hitachi Carbon DDCS. We conducted fixed-point and scanning irradiation experiments at three different energies, both with and without breakpoints. For fixed-point irradiation, we utilized a 2D array detector and an oscilloscope to measure beam intensity over time. The oscilloscope data enabled us to confirm beam-off and beam-on timing due to breakpoints, as well as the relative timing of the RFKO signal, HSST signal, and dose monitor (DM) signals. From these measurements, we analyzed and modelled the temporal characteristics of the beam intensity. We also developed a model for the spot shape and amplitude at isocenter occurring after the beam-off signal which we called flap dose and its dependence on beam intensity. In the case of scanning irradiation, we measured move doses using the 2D array detector and compared these measurements with our model. We observed that the most dominant time variation of the beam intensity was at 1 kHz and its harmonic frequencies. Our findings revealed that the derived beam intensity cannot reach the preset beam intensity when each spot belongs to different breakpoints. The beam-off time due to breakpoints was approximately 100 ms, while the beam rise time and fall time (tdecay ) were remarkably fast, about 10 ms and 0.2 ms, respectively. Moreover, we measured the time lag (tdelay ) of approximately 0.2 ms between the RFKO and HSST signals. Since tdelay ≈ tdecay at HIMAK then the HSST is activated after the residual beam intensity, resulting in essentially zero flap dose at isocenter from the HSST. Our measurements of the move dose demonstrated excellent agreement with the modelled move dose. We conducted the first move dose measurement for a Hitachi Carbon synchrotron, and our findings, considering beam on/off control details, indicate that Hitachi's carbon synchrotron provides a stable beam at HIMAK. Our work suggests that measuring both move dose and flap dose should be part of the commissioning process and possibly using our model in the Treatment Planning System (TPS) for new facilities with treatment delivery control systems with higher beam intensities and faster beam-off control.

Evaluation of Hydraulic Characteristics of Electrohydraulic Proportional Valve (EHPV) for an Auto-Steering Tractor Application

The performance of the electrohydraulic proportional control valve (EHPV) employed in a tractor’s automatic steering system directly influences the steering performance. To develop a highly reliable EHPV, it is essential to analyze the hydraulic characteristics of the EHPV for several working conditions of tractors. This study aimed to measure and analyze the hydraulic characteristics of the EHPV according to tractor working conditions. The flow rate and pressure data of the EHPV were computed through the valve measuring system, and the required power was computed. The experimental conditions were selected based on engine rotational speed and tractor steering angle. As a result, it was discovered that the flow rate, pressure, and power all increased when the engine rotation speed and steering angle conditions increased. Furthermore, the rates of increase in flow rate, pressure, and power based on the increase in the steering angle were higher than when the engine rotation speed increased. In the regression analysis results between the two variables and the hydraulic characteristics of EHPVs, the steering angle demonstrated a higher correlation than the engine rotation speed. In conclusion, the steering angle and engine rotational speed are the major variables in the hydraulic characteristics of EHPVs, and the influence of the steering angle is greater.

Dynamic Drifting Control for General Path Tracking of Autonomous Vehicles

Taking full advantage of tire saturation in high sideslip drifting maneuvers can substantially improve the handling limits of autonomous vehicles. However, most singular motion control methods in conventional drifting controllers have been conducted, which cannot meet general driving needs. In this paper, a hierarchical dynamic drifting controller (HDDC) is designed under both drifting maneuvers and typical cornering maneuvers to uniformly track the general path within and beyond stability limits. Specifically, the HDDC includes the path tracking layer, vehicle motion control layer, and actuator regulating layer. The first layer is designed to achieve accurate path tracking and generate the desired states, and it is algorithmically adaptable and confirmed by MPC and LQR. The second layer is proposed to integrate drifting and typical cornering control by the dynamic drifting inverse model. The third layer can achieve the steering system and wheel speed control to satisfy corresponding commands. Ultimately, the real-time performance of the controller is verified by the ECU hardware-in-the-loop platform. The results with MPC-HDDC and LQR-HDDC show that the HDDC can achieve integrated control of drifting and typical cornering maneuvers and possess high tracking accuracy.

High clearance four-wheel independent electric drive sprayer path tracking control based on self-correction controller

Most traditional agricultural machines use traditional mechanical transmission with fuel drive and front wheel steering, and most of them are manned, with low transmission efficiency, poor flexibility and low level of intelligence. This paper proposes a new high ground clearance sprayer with four-wheel independent drive (4WID) with dual front and rear steering axles. This sprayer is an oil-electric hybrid that not only saves energy, but also reduces the pollution to crops. In this study, a path tracking controller and a self-correcting controller are constructed based on the kinematic model of the sprayer, respectively. The model prediction controller is used as the main controller, and it outputs the desired steering angle and speed of the sprayer based on the current state of the sprayer and the desired path to achieve path tracking control. The self-correcting controller, as an auxiliary controller, adopts fuzzy control method and designs fuzzy control rules based on the driver’s experience to correct the wheel turning angle at the current moment, which not only retains the advantages of model prediction control in linear path tracking, but also improves the control effect in curved path tracking and enhances the path tracking accuracy in large curvature roads. The simulation shows that the maximum lateral deviation of 0.03 m can be achieved by the model predictive control alone, and the maximum lateral deviation is reduced to 0.0141 m after adding the self-correcting controller.

The Relationship Between Mental Workload and Sleep Quantity with Work Fatigue among Haul Dump Truck Operators in Coal Mining

Introduction: In mining production activities, haul dump trucks are utilized for transportation purposes and are driven by an operator. The haul dump truck operator works in a shift work system consisting of 12 working hours. The workers are required to always control the steering speed, causing potential work fatigue. This study aimed to analyze the relationship between mental workload and sleep quantity with work fatigue among haul dump truck operators. Method: This study was a quantitative research study using a cross-sectional design. The population was 179 haul dump trucks operators, while the sample was 123 respondents determined through simple random sampling. The variables of this study were age, years of service, education level, mental workload, sleep quantity, and work fatigue complaints. The data was tested using the Chi-Square test. The instruments used to obtain the data were the National Aeronautics and Space Administration Task Load Index and the Work Fatigue Measurement Tool Questionnaire. Result: The haul dump truck operators (54.5%) had a high mental workload, low sleep quantity (61.8%) with less than 7 hours of sleep, and moderate fatigue complaints (44.7%). There was a significant relationship found between age (P-value = 0.018), years of service (P-value = 0.039), and sleep quantity (P-value = 0.001) with work fatigue complaints. The level of education and mental workload had a significance value above 0.05. Conclusion: Sleep quantity has a significant relationship with work fatigue complaints. This can be responded to by intensifying the information on the socialization of fatigue and the importance of maintaining the quality of sleep.

Research on the Obstacle-Avoidance Steering Control Strategy of Tracked Inspection Robots

Tracked inspection robots possess prominent advantages in dealing with severe environment rescue, safety inspection, and other important tasks, and have been used widely. However, tracked robots are affected by skidding and slipping, so it is difficult to achieve accurate control. For example, the control parameters of a tracked robot are the same during driving, but the pressure, shear force and steering resistance of the robot on the road surface are different, which affects the steering characteristics of the robot on complex terrain. Based on analysis of the structural parameters and steering radius of the robot, the traction force and resistance torque models of the tracked robot were established, and the plane dynamics of the robot’s steering were analyzed and solved. The corresponding relationships between the road parameters, relative steering radius, and lateral relative offset of the robot on three typical roads were obtained. Mathematical models of the robot’s track speed and relative steering radius with and without skid and slip were established. Through simulation analysis of Matlab software, the corresponding relationship between the relative steering radius of the robot and the velocity difference of the two tracks were obtained. Taking angular obstacles as an example, three obstacle-avoidance steering control strategies, once turning in situ center, twice turning in situ center, and large-radius steering were developed. The tracked robot and obstacle multi-body dynamic simulation models were constructed using ADAMS simulation software. The simulation results show that all three methods can complete the steering tasks according to the requirements; however, under the influence of skid and slip, the trajectory of the robot deviates from the ideal trajectory, which has a great impact on large-radius steering, even though the large-radius obstacle-avoidance steering control strategy has the advantages of a smooth trajectory, fast steering speed, and high efficiency. The obstacle-avoidance steering experiments were completed by the robot prototype, which verifies the rationality of robot steering theory, which could provide the corresponding theoretical basis for autonomous obstacle-avoidance navigation control of a tracked robot.

Design of a Liquid‐Driven Laser Scanner with Low Voltage Based on Liquid‐Infused Membrane

Laser energy is commonly used in tissue ablation, wound suturing, and other precise manipulations during surgery. However, currently available laser scanners require further improvements in terms of miniaturization, driving voltage, and stability to steer the laser beam accurately within a constrained environment. Herein, the development of a liquid‐driven laser scanner installed on the end effector of a continuum endoscope to perform fast and reliable laser steering is proposed. The developed laser scanner is 7 mm in diameter and 7 mm in length, and it is actuated with a voltage lower than 15 V due to the liquid‐infused membrane. The miniature size and low driving voltage of the proposed laser scanner facilitate safe laser‐assisted surgery in confined spaces. A theoretical model is established to predict laser spot position quantitatively, and laser steering ability is also tested experimentally. The fiber‐delivered laser beam can be steered for 21.2° (±10.6°) with a standard deviation of 0.3° in 1000 cycles, demonstrating excellent stability. A laser steering speed of up to 27.3 mm s−1 and a reflection loss of less than 3.1% are achieved.

Deep-Learning-Based Network for Lane Following in Autonomous Vehicles

The research field of autonomous self-driving vehicles has recently become increasingly popular. In addition, motion-planning technology is essential for autonomous vehicles because it mitigates the prevailing on-road obstacles. Herein, a deep-learning-network-based architecture that was integrated with VGG16 and the gated recurrent unit (GRU) was applied for lane-following on roads. The normalized input image was fed to the three-layer VGG16 output layer as a pattern and the GRU output layer as the last layer. Next, the processed data were fed to the two fully connected layers, with a dropout layer added in between each layer. Afterward, to evaluate the deep-learning-network-based model, the steering angle and speed from the control task were predicted as output parameters. Experiments were conducted using the a dataset from the Udacity simulator and a real dataset. The results show that the proposed framework remarkably predicted steering angles in different directions. Furthermore, the proposed approach achieved higher mean square errors of 0.0230 and 0.0936 and and inference times of 3–4 and 3 ms. We also implemented our proposed framework on the NVIDIA Jetson embedded platform (Jetson Nano 4 GB) and compared it with the GPU’s computational time. The results revealed that the embedded system took 45–46 s to execute a single epoch in order to predict the steering angle. The results show that the proposed framework generates fruitful and accurate motion planning for lane-following in autonomous driving.

Intelligent Control System of Unmanned Vehicle Based on CAN Controller

In order to enable the driverless vehicle formation controller to automatically control the driving of the fleet, an intelligent driverless vehicle control system based on the CAN controller is proposed. Expansion convolution of different expansion rates is used to obtain multi-scale target information and to fuse the feature information at different scales during upsampling to enrich the semantic information. Finally, the driverless CAN bus communication platform was established, the driverless monitoring interface was developed, and the software program was written; experiments on the steering control, speed control, voltage, current, speed, and angular speed acquisition, respectively, were performed. The experimental results show that the average semantic segmentation accuracy of the obstacles in concentrated vehicles, pedestrians, and bicycles reached 84.6%, and the detection and segmentation accuracy of the models was good. Therefore, the unmanned intelligent vehicle control system designed in this paper can meet the performance requirements of the vehicle control. No matter whether the given desired path is a straight line or a curve, the unmanned car can complete the path tracking control quickly, stably, and accurately.

Ultrafast parallel single-pixel LiDAR with all-optical spectro-temporal encoding

Light detection and ranging (LiDAR) with scanning beam has been widely used in autonomous driving and large-scale manufacturing, where the fast-changing scene is of great interest. However, the acquisition rate of current LiDAR is limited by beam steering speed and the round-trip delay, hindering the development of video-rate LiDAR. Here, we uniquely combine all-optical encoders with wavelength-division multiplexing devices to implement spectro-temporal encoding on the illumination light. Parallel detection and fast spectral scanning can be achieved with only one laser and one single photodetector. Our result shows the fastest single-pixel LiDAR to date, and the detection speed can be multiplied with scalability. We demonstrate 3.56-fold improvement of speed for a maximum of 75 m detection range compared with a serial LiDAR. A record acquisition rate of 21.38 MHz for a maximum of 25 m detection range is obtained. This approach has the potential to break the limitation on current ultrafast single-pixel LiDAR and opens a new paradigm for ultrafast-frame-rate imaging systems.

Multi-beam optical phase array for long-range LiDAR and free-space data communication

We present a multi-beam optical fiber phased array (OPA) with simultaneous light detection and ranging (LiDAR) and free-space optical communication (FSO) data transmission functions. The OPA component can achieve a stable phase and a high turning speed. First, the 55-element OPA is phase-locked through the stochastic parallel gradient descent (SPGD) algorithm to maintain the high light intensity of the main beam in real time. The OPA with a 5 × 11 beam array achieved a wide field of view (FOV) of 58°×32°, a single beam array steering angle of0.48°, and a steering speed of 40 MHz. Subsequently, this OPA was used to detect targets within a range of approximately 2 km, and we theoretically verified that two-dimensional coherent frequency-modulated continuous wave (FMCW) LiDAR could extract the target characteristics. In addition, a free-space optical communication link of up to 3 km was also demonstrated, with a data rate of 8 Gbps. The proposed system easily increases the array-to-MHz steering speed and yields a high-power output.

Wind tunnel studies of wind turbine yaw and speed control effects on the wake trajectory and thrust stabilization

Yaw-based wake steering is pursued as a means to avoid upstream turbine wake influence on downstream turbines. However, the method's effectiveness is likely restricted by rapid wind direction changes that cause wake trajectory oscillations. To address this issue, a thrust-based wake trajectory stabilization method is proposed here. The changes to the cross-stream thrust, resulting from unintentional yaw misalignments, caused due changes in inflow wind direction could be counteracted by changing rotor speed and blade pitch, thus stabilizing wake trajectory oscillations. This paper presents a wind tunnel case study to explore the effect of combining yaw-based wake steering and fast-changing rotor speed control on the cross-stream thrust component and hence on the wake trajectory. Although a feedback controller was not implemented, it is demonstrated that wake trajectory oscillation could be contained using faster actuating mechanisms, such as rotor speed or pitch. Here the wake oscillations due to wind direction changes emulated by sinusoidally yawing the turbine, were restricted by varying turbine rotation speed. Direct thrust measurements as well as detailed Particle Image Velocimetry (PIV) and hotwire wake data show how this combination can successfully suppress the dynamic cross-stream thrust component perturbations and subsequently the wake trajectory oscillation around a desired position.

End-to-end driving model based on deep learning and attention mechanism

End-to-end deep learning has gained considerable interests in autonomous driving vehicles. End-to-end autonomous driving uses the deep convolutional neural network to establish input-to-output mapping. However, existing end-to-end driving models only predict steering angle with front-facing camera data and poorly extract spatial-temporal information. Based on deep learning and attention mechanism, we propose an end-to-end driving model which combines the multi-stream attention module with the multi-stream network. As a multimodal multitask model, the proposed end-to-end driving model not only fully extracts spatial-temporal information from multimodality, but also adopts the multitask learning method with hard parameter sharing to predict the steering angle and speed. Furthermore, the proposed multi-stream attention module predicts the attention weights of streams based on the multimodal feature fusion, which encourages the proposed end-to-end driving model to pay attention to streams that positively impact the prediction result. We demonstrate the efficiency of the proposed driving model on the public Udacity dataset compared to existing models. Experimental results show that the proposed driving model has better performances than other existing methods.

Automated Vehicle Sideslip Angle Estimation Considering Signal Measurement Characteristic

Vehicle slip angle (VSA) estimation is of paramount importance for connected automated vehicle dynamic control, especially in critical lateral driving scenarios. In this paper, a novel kinematic-model-based VSA estimation method is proposed by fusing information from a global navigation satellite system (GNSS) and an inertial measurement unit (IMU). First, to reject the gravity components induced by the vehicle roll and pitch, a vehicle attitude angle observer based on the square-root cubature Kalman filter (SCKF) is designed to estimate the roll and pitch. A novel feedback mechanism based on the vehicle intrinsic information (the steering angle and wheel speed) for the pitch and roll is designed. Then, the integration of the reverse smoothing and grey prediction is adopted to compensate for the cumulative velocity errors during the relatively low sampling interval of the GNSS. Moreover, the GNSS signal delay has been addressed by an estimation-prediction integrated framework. Finally, the results confirm that the proposed method can estimate the VSA under both the slalom and double lane change (DLC) scenarios.