Research On Design Methodology And Key Techniques For Multi-station Fully Automated Balancing Machines

This dissertation combines some content which have been finished for the High Technology R&D Project of China "Research on the Key Technology of High Precision Automatic Correction for Balancing Medium and Small-sized Motor Products " (Granted No.2002AA421110) and the Key Technology Program of Zhejiang Province, China "Research on Key Technologies and Industrialization Development of Automated Balancing Machines" (Granted No. 021101566). Taking high precision, high efficiency, flexible and low cost multi-station fully automated balancing machines as subject investigated, its key technology and computer aided design method was studied against the shortcomings of current domestic balancing machines industry. The contributions are as follows:Firstly, rotor-pedestal system modeling and calibration method of balancing machine was studied. Based on the rigid rotor-pedestal system three-axis dynamic model, the impact of natural frequency and sensitivity of system on performance of balancing machine was quantitatively analyzed, and feasible measure for improving precision of balancing machine was put forth, these give a guide for design and selection rotor-pedestal system, driving system and sensor. Various calibration methods and its system error were analyzed in-depth, an uniform matrix format theoretic formulae of four existing calibration methods—once for all method, influence coefficient method, vector influence coefficient method and multi-linear regression method were presented. The request of four methods for the pedestal system and the measurement system was analyzed, the four methods have different ability of compensating the system error, frequency response characteristics of the pedestal system and the measurement system have a great impact on calibration method, their advantages and disadvantages are clarified analytically, and once for all calibration method based on rotational speed learning was put forth, which has higher precision than traditional once for all method.Secondly, digital signal processing method of measurement system of balancing machines was studied in detail. Several methods extracting magnitude and phase of unbalance signal were presented and clarified analytically, the significance of integral periodic signal sampling was indicated, and error of these method was analyzed. Precision of balancing machines lies on the ability of extracting small signals buried in background noise, transfer functions of six digital filters—synchronous time average, Chebyshev narrow band-pass filter, integral digital filter, digital tracking filter, adaptive filter and harmonic wavelet filter were given, their applications in measurement system of balancing machines were investigated in detail. From the point of view of practicability, their filtering abilities and realization methods were analyzed and compared. As a result, some instructive conclusions are provided for selection of digital filter of balancing machines. In addition, the influence of speed fluctuation on precision of balancing machines was quantitatively analyzed, and the demand of speed stabilization was indicated for high precision balancing machine.