融合神经肌肉动力学和QN认知体系的驾驶员车辆控制模型

发布时间：2018-12-11 23:50

【摘要】：驾驶员车辆控制模型不但可以揭示驾驶员的驾驶机理，而且可以计算、仿真和预测驾驶员的车辆控制行为，帮助改进车辆和辅助驾驶系统的设计过程。此外，它们可以为智能辅助驾驶系统和无人驾驶车辆控制技术的研究提供新的思路。因此，驾驶员车辆控制模型的研究有重要的科学意义和实用价值。现有的研究或从控制（包括驾驶员神经肌肉系统方面）或从认知角度发展驾驶员车辆控制模型。因此，这些模型不能全面的揭示驾驶员的控制机理和仿真驾驶员的车辆控制行为，这也限制了它们在帮助发展辅助驾驶系统等方面的价值。为了解决上述问题，本文建立了融合神经肌肉动力学和QN认知体系的驾驶员车辆侧向控制模型，，提出了运动辅助任务条件下的驾驶员车辆侧向控制的双任务建模方法。该模型以QN认知体系为框架可以反映驾驶员认知能力和局限，与控制理论的结合可以实现车辆控制的数学描述，融入神经肌肉动力学可以体现驾驶员的神经肌肉系统和车辆转向系统之间的动态交互。本论文取得的主要创新性成果： 1）改进了基于QN认知体系的驾驶员模型，提出了表征驾驶转向控制的神经肌肉动力学模型，在此基础上，通过融合神经肌肉动力学和QN认知体系模型，建立了一种新颖的驾驶员车辆侧向控制模型。 2）利用建立的驾驶员模型，揭示了神经肌肉动力学参数（包括肌肉协同收缩刚度和反射控制增益）对驾驶员侧向控制性能的影响规律。 3）在建立的驾驶员模型的基础上，通过应用多任务调度方法，提出了一种运动辅助任务条件下的驾驶员车辆侧向控制的双任务建模方法。这些创新成果的取得不仅可以帮助更好地揭示驾驶员的车辆控制机理，解释运动辅助任务对驾驶主任务的影响，也可以为智能车辆和智能辅助驾驶系统的研究和开发提供支持。
[Abstract]:The driver's vehicle control model can not only reveal the driver's driving mechanism, but also can calculate, simulate and predict the driver's vehicle control behavior, and help to improve the design process of vehicle and auxiliary driving system. In addition, they can provide new ideas for the research of intelligent auxiliary driving system and driverless vehicle control technology. Therefore, the study of driver's vehicle control model has important scientific significance and practical value. Existing studies have developed driver's vehicle control models either in terms of control (including driver's neuromuscular systems) or from a cognitive perspective. Therefore, these models can not fully reveal the driver's control mechanism and the simulation of the driver's vehicle control behavior, which limits their value in helping to develop the auxiliary driving system. In order to solve the above problems, a driver's lateral control model based on neuromuscular dynamics and QN cognitive system is established, and a two-task modeling method for driver's lateral control under the condition of motion assistant task is proposed. The model based on QN cognitive system can reflect the cognitive ability and limitation of drivers, and the mathematical description of vehicle control can be realized by combining with control theory. The integration of neuromuscular dynamics can reflect the dynamic interaction between the driver's neuromuscular system and the vehicle steering system. The main innovative achievements in this thesis are as follows: 1) the driver model based on QN cognitive system is improved, and the neuromuscular dynamics model which represents the steering control is proposed. By combining neuromuscular dynamics with QN cognitive system model, a novel driver lateral control model was established. 2) the influence of neuromuscular dynamic parameters (including muscle cocontraction stiffness and reflex control gain) on driver's lateral control performance is revealed by using the driver's model. 3) based on the established driver model, a two-task modeling method for driver lateral control under the condition of motion assistant task is proposed by applying the multi-task scheduling method. The achievement of these innovations can not only help to better reveal the mechanism of driver's vehicle control, explain the influence of motion assistant task on driving main task, but also provide support for the research and development of intelligent vehicle and intelligent auxiliary driving system.
【学位授予单位】：北京理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U491.25

【参考文献】