The application belongs to the technical field of intelligent
medical treatment, and provides an orthopedic reduction
robot force-position sensing self-adaptive regulation method. The method constructs a three-dimensional model of a diseased side bone containing a
bone strength parameter, a healthy side mirror model, and a personalized
soft tissue mechanics model containing muscles and ligaments through preoperative CT and MRI multi-
modal images. In combination with an expert reduction template and a
reinforcement learning algorithm, an initial reduction path and parameters are generated in combination with force and position as dual targets, and a full closed-
loop control is formed through an adaptive regulation network by fusing real-time force and position data with preoperative
mechanics parameters, so that the upper limit of mechanical arm movement and force control is dynamically adjusted, reduction is completed when a preset accuracy is reached, automatic connection from
preoperative planning to intraoperative execution is realized, and the mechanical arm can independently complete the reduction operation according to real-time regulation instructions without frequent manual path correction by the doctor.