This study aims to create a system that can be used to evaluate vehicle characteristics while simultaneously controlling human body behavior through numerical simulations. The proposed system consists of a vehicle model, a human body dynamics model, and a musculoskeletal model. In the present paper, a human body dynamics model using multibody dynamics is proposed. However, attempting to implement a whole-body model would necessitate dealing with multiple degrees of freedom and give rise to

more »

... ematic phenomena. Furthermore, the influences of human motion are uncertain and difficult to parameterize. Accordingly, in the present research, the human model is limited to the head and trunk of a human body riding inside a vehicle. This human body dynamics model is composed of an internal model and an external model. The internal model incorporates a motion control model. The internal model, which is composed of an inverse model and a forward model, generates commands to control body motion, while the external model simulates the actual body motion. Then, in order to identify the parameters of the motion control model, the motion of maintaining posture is measured using a simple experimental device that can simulate horizontal acceleration applied to a subject. In order to demonstrate the effectiveness of the proposed human body dynamics model, a simple human dummy model (which simulates the experimental model used for experiments such as automobile collisions) that consists of only a spring and a damper was created. Comparing this dummy model with the human body dynamics model reveals that the human body dynamics model can simulate details of human motion that the simple dummy model cannot. and the recruitment of test subjects is costly and time consuming. On the other hand, as computer performance has improved rapidly over the years, computer simulation technology has become more sophisticated. Research has been conducted using numerical simulation models, including motion analysis considering the coupled passenger-car body vibration (Tohtake et al, 2006) and vibration and ride comfort analysis using passenger-seat dynamics models (Saito et al, 2013) . Moreover, a detailed human motion numerical simulation model, called the Digital Human (Hase and Yamazaki, 2002) , is currently being developed. The Digital Human has been used to analyze automobile collisions and human walking. When these simulation methods are applied to comfort research, in order to quantitatively evaluate the loads felt by human passengers, the passive properties and motion control of human passengers subjected to vibrations due to external forces must be clarified. In the present study, a simulation system that can consistently calculate various motions of a vehicle and its passenger load is proposed. The proposed system consists of a vehicle model, human model, and musculo-skeletal model, each of which can be modified independently. In the proposed system, the human model has 43 degrees of freedom, and the effects of the parameters for each human joint remain unclear. Moreover, setting exact values of these parameters is difficult. Therefore, we herein focus on motion control of the human head and trunk because this part of the body accounts for a large portion of the overall human body. The simplified human model that can identify characteristics of motion control by actual human response is proposed. Personal quality that individuals have can be studied using this model. A simple head and trunk model is formulated with multibody dynamics that has high extensibility, which enables future extension to the whole-body model. Then knowledge of motion control using simplified human model can be used for motion control of the whole-body model. Moreover, the seat model and the contact situation between the seat and the human body can be defined in detail. First, the controlling parameters of the model are set through simple experiments, and the effectiveness of the suggested model is demonstrated by comparing the response of the motion control model with the parameters obtained using a spring-damper model. This paper is based upon , but the current paper includes the following additional research: detail of modeling human body dynamic model and identification of human body dynamic model.