摘要

To overcome the shortcomings of the traditional free-form surface tracking measurement data sampling methods, a new data accept and reject method which combines isochronous indistinctive sampling with off-line equal-error accurate arithmetic is designed. Dual-thread and dual-buffer dynamic storage technique is adopted to solve the problem of mass data information storage, and data acquisition processing is divided into two links: data sampling thread and data storage thread, each link being achieved by an individual thread and can concurrently assess data buffer without overlapping. SQLite3 is chosen to realize data storage, and the storage speed is tested by experiments. To reflect the geometric characteristic of the model surface figure accurately and make the measured data close to the real model curve more, the circle ratio method is utilized to achieve the curve smoothing treatment. The designed data sampling method enhances the sampling precision and sampling efficiency. Finally, a bicycle saddle model intersection's self-learning control process is simulated with Matlab. Simulation results demonstrate the workability and the effectiveness of the proposed method.