bob手机在线登陆学报自然版

cover

2022, 42(7)

Abstract(50) PDF(3)

Abstract:

2022, 42(7): 1-2.

Abstract(91) HTML(57) PDF(16)

Abstract:

Review on Machine Learning Methods for Motion Planning and Control Policy of Intelligent Vehicles

GONG Jianwei, GONG Cheng, LIN Yunlong, LI Zirui, LÜ Chao

2022, 42(7): 665-674. doi:10.15918/j.tbit1001-0645.2022.095

Abstract(501) HTML(135) PDF(139)

Abstract:
Intelligent vehicles have achieved a considerable development in technologies and can fulfill the basic functions of autonomous driving in a limited closed environment. However, results of actual road tests show that the current technologies of intelligent vehicles still have many limitations and their large-scale application in complex urban and off-road environments still faces many challenges. As one of the key technologies, the motion planning and control technology has basically formed a complete theoretical system and has been widely applied in engineering. However, the traditional methods still have some defects in practical application, such as the inability of understanding dynamic and complex scenes, poor adaptability for different scenes, high complexity of the model, and difficulty in parameter tuning. Due to the strong ability in knowledge representation and model fitting, machine learning methods have been widely applied in perception and navigation technology for intelligent vehicles. In order to solve the problems of generalization and applicability in traditional motion planning and control techniques, many researchers have also devoted themselves to exploring the usage of deep learning, reinforcement learning, and so on machine learning methods in motion planning and control policy for intelligent vehicles. In this paper, machine learning-based methods were reviewed for motion planning and control in intelligent vehicles, analyzing the existing policy learning methods for motion planning and control from three aspects, including basic framework, basic learning paradigms, and different planning and control methods based on learning. Finally, the research status and future development directions were summarized and prospected.

Fuel Economy of Lane Changing Trajectory for Intelligent Vehicle

JIN Hui, ZHAO Yanjie, GUANG Hao

2022, 42(7): 675-681. doi:10.15918/j.tbit1001-0645.2021.250

Abstract(183) HTML(58) PDF(27)

Abstract:
In this article, the fuel economy of lane changing trajectories was studied based on the space and trajectory curvature constraints of intelligent vehicle. Based on an engine transient fuel consumption model, the most economical speed for driving on the flat roads was calculated, and the third-order Bezier lane changing trajectory model was established for various constraints. The results of Matlab/Simulink and Carsim co-simulation verification show that, the fuel consumption of the Bezier lane changing trajectory developed in the paper is 3.49% and 0.77% less than the existing sin-tanh andx-sin lane changing trajectories, and its maximum lateral acceleration value is also 31.75% and 7.45% lower than sin-tanh andx-sin lane changing trajectories respectively. Therefore, on the basis of ensuring the safe and comfortable driving of intelligent vehicles, the fuel consumption of vehicle is the least by Bezier lane changing trajectory, and the corresponding maximum lateral acceleration is also the smallest. All those prove the superiority of the Bezier lane-changing trajectory.

Trajectory Planning of Dynamic Take-off and Landing of Deformable Aerial-Ground Platform

WEI Wei, FAN Kangdi, WANG Ruolin, ZHAO Molei, ZHAO Zhiran, WANG Jianfeng

2022, 42(7): 682-687. doi:10.15918/j.tbit1001-0645.2021.325

Abstract(179) HTML(66) PDF(20)

Abstract:
Aerial-ground platform has multi-domain maneuverability and can adapt to various complex environments through the conversion of land and air mode, but the land and air mode conversion is mostly static take-off or hovering descent, which is not conducive to the full display of the maneuverability of the aerial-ground platform. Aiming at a deformable aerial-ground amphibious platform with deflectable power mechanism, the flight dynamics model of the aerial-ground platform was established based on the Newton-Euler equation, the time sequence of the deflection angle was planned to obtain dynamic constraints, and the relative time optimal objective function was determined. The fifth-order polynomial was used to fit the two-dimensional trajectory, and the trajectory tracking controller was designed according to the PID control method, and the trajectory planning and control simulation were carried out. The results show that the dynamic switching time is shortened by 23.02% compared with the static switching time, the dynamic switching planning trajectory is smooth, there is no overshoot in the altitude direction, and the controller can better track the target flight trajectory.

Graph Representation Method for Pedestrian Intention Recognition of Intelligent Vehicle

LÜ Chao, CUI Gege, MENG Xianghao, LU Junyan, XU Youzhi, GONG Jianwei

2022, 42(7): 688-695. doi:10.15918/j.tbit1001-0645.2021.330

Abstract(234) HTML(106) PDF(37)

Abstract:
The problem of pedestrian-vehicle conflict in intelligent driving scenes is closely related to pedestrian crossing behavior. In order to enable advanced driving assistance system (ADAS) to have the function of identifying pedestrian crossing intentions and raising advanced warning of pedestrian-vehicle collision events, a pedestrian crossing intention recognition framework based on graph representation learning (GRL) method is proposed. It uses open source tools to generate pedestrian skeleton information. Then it establishes a graph model to represent the characteristics of pedestrian action sequence by taking the skeleton key points of each frame of pedestrian within a sequence as nodes, as well as taking the natural connections, the topological correlations and time-domain relationships between skeleton joints as edges. Taking the graph structure data as the input, the pedestrian crossing intention recognition model is trained based on support vector machine (SVM). The results show that the classification accuracy of pedestrian crossing intention can reach 90.29%. The proposed method can effectively identify the pedestrian crossing intention, which is of great significance to improve the safety of intelligent vehicle decision-making.

Local Dynamic Obstacle Avoidance Path Planning Algorithm for Unmanned Vehicles Based on Potential Field Method

ZHAI Li, ZHANG Xueying, ZHANG Xian, WANG Chengping

2022, 42(7): 696-705. doi:10.15918/j.tbit1001-0645.2021.333

Abstract(326) HTML(487) PDF(105)

Abstract:
To achieve dynamic real-time obstacle avoidance of unmanned vehicles, a local obstacle avoidance path planning algorithm was proposed based on artificial potential field method. Firstly, improving the potential field environment and the potential field force were arranged in the new method to solve the local minimum value and target unreachable problem of the traditional potential field method. And then, considering the safety of vehicle collisions, the working conditions of lateral dynamic obstacles and the same direction dynamic obstacles were analyzed, and a dynamic window method was used for real-time dynamic obstacle avoidance planning. To ensure path flatness and traceability, a BSL curve was used to smoothing the planned path. Finally, the proposed control algorithm was verified under the co-simulation platform of CarSim and Matlab/Simulink. The simulation results show the effectiveness, safety and traceability of the planning algorithm for obstacle avoidance.

Dual Closed-Loop Control of Intelligent Vehicles Trajectory Tracking Based on Steering Response Characteristics

XU Bin, ZHANG Dapeng, YANG Haiyang, HAN Ziyong

2022, 42(7): 706-712. doi:10.15918/j.tbit1001-0645.2021.328

Abstract(219) HTML(106) PDF(52)

Abstract:
The accuracy and robustness of intelligent vehicle trajectory tracking are important characteristic of vehicle motion control performance. Many Researches of tracking control based on path preview information have significantly improved vehicle performance. However, the insufficient response of the vehicle steering system have caused many difficulties for the vehicle to accurately track the reference trajectory based on the preview information in real time. To solve this problem, introducing state of the steering system in real time, a dual closed-loop trajectory tracking control structure was established to ensure the robustness of the intelligent vehicle trajectory tracking control algorithm for the insufficient response of the steering system. Concretely, in the outer loop of this structure, a model predictive control method was used to solve the optimal steering angle based on preview information. And in the inner loop of this structure, a PID method was used to design a feedback control law based on steering state error to compensate for insufficient steering response. Coupling control input, the dual closed-loop structure can ensure the robust optimal tracking control of the vehicle. Finally, the effectiveness of the dual closed-loop control structure was verified in Carsim and Simulink co-simulation.

Vehicle Mass-Centroid Sideslip Angle Estimation Based on Extension Fusion of Fuzzy Sliding-Mode Observer and Sensor Signal Integral

WANG Hongbo, XU Shihan, ZHOU Daolin, WANG Xiangyu, LIU Xinyu

2022, 42(7): 713-722. doi:10.15918/j.tbit1001-0645.2021.300

Abstract(176) HTML(79) PDF(24)

Abstract:
Vehicle mass-centroid sideslip angle plays an important role in the judgment of vehicle lateral stability. Overestimation or underestimation of the sideslip angle will affect the stability control system. At present, the estimation of sideslip angle still has a large error and has not been practically used in engineering. In order to reduce the observation error and improve the practicability of the estimation system, a robust fuzzy second-order sliding-mode observer was proposed to calculate the observed value of the sideslip angle, and the integral value of the sideslip angle was calculated by using the inertial measurement unit signal. Then, the advantages and disadvantages of the two estimation methods were analyzed, and the observation estimation value and the integral estimation value of the sideslip angle were extensively fused to realize the correction of the observation value with the sensor signal. Finally, the estimation method of sideslip angle was verified by Simulink/TruckSim simulation and hardware in the loop simulation. The effectiveness of the proposed method is proved based on the control effect in real vehicle test under the constant circular acceleration condition. The results show that the proposed method can accurately reflect the actual sideslip angle, and has good reliability and practicability.

Fault Estimation Based Fault-Tolerant Control Method for Unmanned Vehicle

XIA Qiu, CHEN Long, XU Xing, CAI Yingfeng, CHEN Te, WU Xiaohan

2022, 42(7): 723-732. doi:10.15918/j.tbit1001-0645.2021.161

Abstract(164) HTML(65) PDF(22)

Abstract:
In order to improve the reliability of unmanned-vehicle control system, a fault-tolerant control method was proposed for unmanned vehicle path following, considering the influence of both steering control system fault and unknown disturbance on path-tracking effect of unmanned vehicle. Firstly, analyzing the input fault of steering control system and characterizing nominal fault from the unknown system interference, a mathematical model was established. And then, a differential equation of nominal fault was constructed with a high-order sliding mode observer, and the estimation method of vehicle sideslip angle and nominal fault was designed based on adaptive cubature Kalman filter, providing reliable information source for fault-tolerant control of unmanned vehicle. Finally, a fault-tolerant controller was designed based on sliding mode control method for path following of unmanned vehicle to prove its convergence. The results of co-simulation and hardware in the loop test show that, the proposed estimation method can get accurate and reliable estimation results of vehicle sideslip angle and nominal fault, and compared with no fault-tolerant control, the designed fault-tolerant path following controller can significantly improve the control performance of vehicle in the face of fault and interference, and at the same time ensure the path following ability and its own stability of vehicle.

Numerical Simulation Research on Impact Failure and Damage Evolution of Cemented Backfill

LI Xianglong, LI Qiang, WANG Jianguo, YANG Changhui, TAO Zihao, ZUO Ting, ZHAO Zehu

2022, 42(7): 733-740. doi:10.15918/j.tbit1001-0645.2021.189

Abstract(138) HTML(51) PDF(28)

Abstract:
In order to quantitatively describe the damage degree and failure process of the cemented backfill under dynamic load, numerical simulation software was used to perform SHPB dynamic impact on the cemented backfill, and the feasibility of the numerical simulation method was verified by the indoor SHPB impact test results. For 4 kinds of cemented fillings (with cement-sand mass ratios of 1∶4, 1∶6, 1∶8 and 1∶10 respectively) made under different impact speeds (1.5, 1.7, 1.8, and 2.0 m/s), the micro-crack density method was used to define the damage variable valued, and a numerical simulation study of the damage law and the failure process was conducted. The results are as follows. The wave shaper can be used in the numerical simulation to obtain a more ideal rectangle wave, making the stress on the same plane element of the specimen uniform without stress concentration. The numerical simulation results show the dynamic failure process of the cemented filling body, and the overall failure trend is that the edge peels off and the crack extends to the inside. In the process of increasing the loading speed from 1.7 m/s to 1.8 m/s, the damage variable increases by more than 10%; during the process of increasing the impact speed from 1.5 m/s to 2.0 m/s, the variation ranges of the damage variabledof the cemented filling body with the cement-sand mass ratios of 1∶4, 1∶6, 1∶8 and 1∶10 are 0.238～0.336, 0.274～0.413, 0.391～0.547, and 0.473～0.617, respectively. When the lime-to-sand ratio changes from 1:6 to 1:8, the damage “jumps” remarkably.

Performance Analysis and Structure Optimization of Synchronous Switched Reluctance Motor

DONG Lei, RAN Maoying, SHAO Liwei, GAO Mengqi, YU Kunyang

2022, 42(7): 741-748. doi:10.15918/j.tbit1001-0645.2021.172

Abstract(269) HTML(96) PDF(23)

Abstract:
In order to solve the problem of complex rotor structure, deficient mechanical strength, and difficulty in high-speed applications of synchronous reluctance motor, a new type of synchronous switched reluctance motor (SSRM) was proposed. Taking the both advantages of switched reluctance motor and a synchronous reluctance motor, the new motor rotor was designed with several excellences including simple manufacturing process for punching sheet, strong reliability, suitable for high-speed fields, high salient pole rate and strong torque output capability. Based on the electromagnetic design of the new SSRM, the parameters of the motor were optimized by the orthogonal method, taking the average torque, torque ripple and motor efficiency as the optimized performance indicators. The finite element simulation and experiment results show that the optimized SSRM possesses the characteristics of low torque ripple and high efficiency.

A Micro-Expression Apex Frame Spotting Method Based on Optical-Flow-Dual-Input Network

ZHENG Shuhua, CHEN Mengxin, WANG Xiangzhou, GONG Xueya

2022, 42(7): 749-754. doi:10.15918/j.tbit1001-0645.2021.135

Abstract(184) HTML(76) PDF(23)

Abstract:
Micro-expression apex frame contains abundant micro-expression information. In order to spot the apex frame accurately, a neural network classification was proposed based on optical flow characteristics. Taking prior knowledge as rules, a detection method was designed to realize micro-expression apex frame spotting. Firstly, optical flow information was extracted from the image in a fixed size sliding window. And then, the spatial and temporal features of optical flow information in x and y directions was extracted and classified based on dual input network. Finally, according to the trade-off rules based on prior knowledge of micro expression, a post-processing was carried out to improve the detection accuracy. The experimental results on data set CASMEⅡtesting show that the apex spotting rate (ASR) andF₁-score can reach up to 0.945 and 0.925 respectively.

Influence of Dual Stealth Aircraft Coherent Jamming on Monopulse Radar and Impact Effectiveness Study

WANG Chunyang, LIU Mingjie, ZENG Huiyong, ZHOU Changlin

2022, 42(7): 755-763. doi:10.15918/j.tbit1001-0645.2021.185

Abstract(176) HTML(107) PDF(11)

Abstract:
Due to the insufficient analysis on the specific impact of dual stealth aircraft formation self-defense coherent jamming on monopulse radar and the lack of rational evaluation to error effectiveness, considering the random characteristics of the phase difference between the dual aircraft echoes, a dual stealth aircraft coherent jamming model and an angle error evaluation model were established. Setting the track of dual aircraft and extracting the dynamic RCS of dual aircraft, a calculation formula was deduced for the random angle error caused by the dual aircraft coherent jamming under the echo. And then several processes were carried out including solving the first order digital characteristics of random angle errors, establishing the angle error validity evaluation criterion according to the miss distance, and calculating the validity probability of angle error. Simulation results show that the angle error caused by the dual stealth aircraft coherent jamming is relatively larger and the random fluctuation is weak. The stable jamming effect and large error validity probability to radar can insure the security of formation penetration.

Dynamic Response of E36 Shipbuilding Steel Under High Pressure Impact

ZHAO Pengduo, HUO Guojing, WANG Qi, ZHANG Lei, LI Mao, LI Xianyu, ZHANG Zhaohui, ZHANG Shunzhong, HE Jianye

2022, 42(7): 764-772. doi:10.15918/j.tbit1001-0645.2021.188

Abstract(164) HTML(62) PDF(47)

Abstract:
To indicate clearly the dynamic response mechanism and deformation strengthening mechanism of shipbuilding steel materials under ultra-high strain rate, the free surface velocity-time curve of E36 shipbuilding steel under impact pressure of 10, 20 and 30 GPa was measured through one-dimensional plate impact test, and the Hugoniot elastic limit and spall strength of E36 shipbuilding steel were calculated. ANSYS software was used to simulate the temperature field under different impact pressure. The damage evolution law and deformation strengthening mechanism of E36 shipbuilding steel under high-pressure impact were studied based on SEM, TEM and other techniques. The results show that the spalling occurs in the materials under the above mentioned impact pressures, and the damage mechanism is the nucleation, growth and aggregation of micropores and microcracks. With the increase of impact pressure, the Hugoniot elastic limit of E36 shipbuilding steel change little while the spalling strength gradually increases. Phase transformation strengthening, dislocation strengthening and twin strengthening are the main strengthening mechanisms of E36 shipbuilding steel under high pressure and high strain rate.

Submission System

Author Related

Writing Norms

Links

2022 Vol. 42, No. 7

Archive