Hui Yan Chen

A Cost Function-Oriented Quantitative Evaluation Method for Unmanned Ground Vehicles

Task driven approach is widely used as an evaluation method for intelligent vehicle, however this approach may lead to many teams using the conservative approach to complete the task. Although the competition task can be completed, it has deviated from the goal of technological development actually. A cost function-oriented quantitative evaluation method is proposed in this study. The time to complete each task and the quality of each indicator are considered in the evaluation method. The cost function-oriented quantitative evaluation method guides the intelligent vehicle’s development in the low-cost index (i.e. high technology) direction. The evaluation results in the 2010 Future Challenge: Intelligent Vehicles and Beyond (FC’ 2010) competition showed that the proposed method can quantitatively evaluate the overall technical performance and individual technical performance of unmanned vehicles.

Design on Hierarchical Testing System for Unmanned Ground Vehicles

The test system for technical abilities of unmanned vehicles is gradually developed from the single test to comprehensive test. The pre-established test and evaluation system can promote the development of unmanned ground vehicles. The 2009 Future Challenge: Intelligent Vehicles and Beyond (FC’09) pushed Chinas unmanned vehicles out of laboratories. This paper proposed to design a more scientific and comprehensive test system for future competitions to better guide and regulate the development of Chinas unmanned vehicles. According to the design idea of stage by stage and level by level, the hierarchical test content from simple to advanced, from local to overall is designed. Then the hierarchic test environment is established according to the levels of test content. The test method based on multi-platform and multi-sensor is put forward to ensure the accuracy of test results. The testing criterion framework is set up to regulate future unmanned vehicle contests and to assess the unmanned vehicles scientifically and accurately.

Comparison of Single-Layer and Multi-Layer Laser Scanners for Measuring Characteristic

This paper presents our initial studies on comparing properties of two laser scanners. The research mainly consists of two parts: tests on measurement characteristics of laser scanners and performances of them under different environments. Through the tests on measurement characteristics, we deeply understand the measurement drift over time and measuring stability of the two scanners. From comparing their performances under different environments, we found the suited application environment for the two scanners. The conclusion from the whole research provides a support for subsequent studies on intelligent vehicles.

Teleoperation Transparency Using Model Predictive Control

Teleoperation means operating a vehicle or a system remotely over a distance.Teleoperation extends human capabilities to perform tasks remotely by providing the operator withsimilar conditions as those at the remote location. Generally, operator should be supported inreal-time with an accurate data about the teleoperation environment. This paper proposes atime-delay compensation algorithm for environment construction to modify the well-developed BITAGV to be teleoperated. This Algorithm includes predicting the vehicle position in the future timeusing Model Predictive control (MPC). Then, environment is constructed according to this positionusing the Laser scanner. Real and simulation experiments are presented to illustrate the performanceand effectiveness of the algorithm in compensating time-delay for Teleoperation.

Study on Data Acquisition for Heavy Off-Road Vehicle AT Testing

Based on the analysis of the testing automatic transmission (AT), the AT’s solenoid valve control signals conditioning circuits and revolving speed F/V converter conditioning circuits are designed, and the data acquisition system is designed based on virtual instrument. The system is composed of the PCI-6024E data acquisition card, signal conditioning circuits and computer, and the acquisition procedure is programmed in graphical language LabVIEW. Finally, the data acquisition system has been verified on vehicle test bench which loads ZF-Ecomat 6HP500 AT. Controlling signals of switching solenoid valves, duty cycle signal of proportional solenoid valve, accelerator aperture signal, oil pressure signals and speed signals are collected, it turns out that the data acquisition system has good feasibility.

The Power System Active-Synchronizing Control of the PHEV during the AMT Shifting Process

For the parallel hybrid electrical vehicle (PHEV) with motor equipped in front of the transmission, the inertia of the transmission-input increases with the power system and the clutch assembly. To reduce the synchronizer slipping friction power, the input speed of the transmission should be controlled, so the active-synchronizing control is applied. To enhance the speed-adjusting precision and shorten the synchronizing time, engine-motor joint speed-adjusting-control is utilized. Finally, a vehicle simulation is carried out, and the test result proves that the method is feasible.

Researching on Clutch Rotational Speed Signal Processing of AMT Vehicle

Aimed at clutch magnetoelectric tachometric transducer, rotational speed signal processing had achieved good results in hardware circuit and software algorithm two aspects. Firstly in hardware aspect RC Low-Pass Filter’s parameters are optimized by frequency-domain analysis of actual rotational speed signal according to vehicle travel features. As a result, useful information is reserved and high-frequency noise is shielded effectively. Secondly in software aspect, through analyzing the clutch control system’s characteristics and speed data’s structure, explicit Euler formula is adopted to predict the data’s structure. Maximum likelihood estimate of Mathematical Statistic Analysis is used to raise the data’s utilization and enhance the data’s robustness in order to provide the effective and stable parameters for clutch control.

Design on the Electromechanical Shift Mechanism of AMT

According to the performance characteristic of the shift system of AMT, the design requirements of electromechanical shift mechanism are developed, and the electromechanical shift mechanism of AMT for the off-road vehicle is designed. Based on the shift force and shift time, parameters of selection and shift operating mechanism are designed. The mechanism is analyzed by FEA software, affording theoretical principle for the structure improvement and its final design. The test result proves that the electromechanical shift mechanism is reasonable and feasible of the application of AMT system.

PHEV Torque Control of the Pre-Shift Process Based on ANFIS

To make the parallel hybrid electrical vehicle (PHEV) equipped with automatic mechanical transmission (AMT) shift without disengaging clutch, the speed and the torque of the engine and the motor must be controlled precisely and reasonably. Because the engine and the motor all equipped before the AMT, the input shaft moment of inertia of the transmission is great. To ensure the synchronizer of the transmission disengage with a small force and reduce the tooth wear of the synchronizer, we must control the torque of the motor based on different conditions to make the AMT work with small torque transmission before shifting. In view of the uncertainty and complexity of the system, we use ANFIS method to determine the target torque of the motor to make sure the synchronizer disengages with the transmission transferring little torque. Finally, we make a vehicle simulation carried out by the software “cruise” and prove that the method is feasible.

Longitudinal Dynamic Model of AMT Vehicle for Following Condition

As an important part of vehicle following control system, vehicle longitudinal dynamic model had significant effect on the research and development phases. The clutch engaging and exhaust brake process of AMT vehicle represented strong nonlinearity, and the theoretical models were complex and depended on many parameters. Focusing on the requirement of following control system, the models of the two processes above were simplified for the engineering application by combining theoretical analysis and real vehicle test data to describe the dynamic processes. Finally, the vehicle model based on MATLAB was build and the comparative experiments were conducted. The comparative results show that the simulation results agree well with the real vehicle tests.