Research On The Model For Identifying Grating Displacement Sensor‘s Error Characteristic Parameters And Algorithm For Real-time Error Self-correcting

Grating electromagnetic induction displacement sensor like inductosyn or rotary transformer has been widely applied in high-precision CNC machine tools and military fields because of its advantages of high detection precision, strong anti-interference performance and low cost. Presently, for this kind of grating displacement sensor, an important way to improve its accuracy is to increase the groove density and precision of the grating lines, but it becomes even more difficult. The method of error separation and correction has become the alternative way to improve the measurement quality of the sensor of this kind.Many error separation and correction methods can only passively and mechanically track and correct the errors from a mathematical point of view, while not combine the mechanical structure and the electrical characteristics of the sensor itself well. Thus these methods cannot fundamentally trace to the errors’ sources. On the other hand, the calibration of the sensor system in the dynamic measurement process must rely on the higher precision standard instrument. The off-line error correction data will be obtained in the process and the data will be used to calibrate the error of the on-line measurement system. Therefore, these methods are of great limitation. In this paper, funded by a National Natural Science Foundation of China(No:51275551), the research of the multi-pole grating electromagnetic induction displacement sensor based on the above problems will be focused.(1) The basic structure and the measurement theories of the grating displacement sensor system combined with the mechanical structure and the electrical characteristics of the sensor itself will be discussed. The formation factors, the impact pattern and the exploration of the calibration of the main errors such as the quadrature error, the amplitude error, the phase shift error, the function error, the random noise error and so on will also be elaborated.(2) Establish a model and then present simulation analysis according to some factors of the errors to reveal the influence rules of the measurement system. A mathematical model of the representation error characteristics set up by the perspective of the errors transmitting with the signal processing flow, a signal processing proposal of error separation and calibration inside a piece of all-in-one DSP processor integrated displacement calculation has been come up with to pre-calibrate some of the errors. The theoretical derivation has been carried upon this proposal, the experimental system hardware has been made and the test platform has been set up to verify the error pre-calibration experimental proposal.(3) Combined the error-calibration method discussed in(2), a new error self-calibration measurement system structure with coaxial multiple sensors will be put forward to construct the error parameters identification mathematical model in the case of using no benchmark instrument. On principle of system identification theory, the error model parameters have been used as the real-time parameters to proceed recurrence identification, the real-time identification of the sensor error characteristics parameters has been realized, thus the real-time separation and self-calibration of the errors have been achieved according to the sample sequences of the measurement system.(4) An electromagnetic simulation analysis has been carried on to the sensor firstly, then an error self-calibration experimental platform has been built and a data acquisition and self-verification algorithms software has been programmed. Thus contrastive experiments can be achieved under different parameters for improving the model and the self-calibration algorithms and also the problems encountered in the algorithms experiments has been explored.The on-line error processing model and the error self-calibration algorithm established in this subject have been successfully applied in the practical experiment, achieving the error self-calibration of the measurement system without the usage of the benchmark instrument meanwhile improving the real-time measurement accuracy and stability of the system.