Reliability Analysis on the Speed Measurement Redundant System of AMB Based on Markov Process
Yan Zhou,ZhenGang Shi,Ni Mo,GuoJun Yang,Zhe Sun
2014-01-01
Abstract:This article uses the Markov process to analyze the speed measurement redundant units on the active magnetic bearings (AMB) system, especially on the aspects of the reliability, redundancy and maintenance issues. The reliability mathematical model is established, and the state transition diagram and the state transition matrix are deduced under the different conditions, furthermore the expression on reliability of the speed redundant is given. According to the different failure rate of single speed sensor, the curve about the measurement unit reliability versus time is calculated. The methods mentioned above provides a new and practical approach to improve the reliability of the system and to optimize the speed measurement redundant unit, which also can to be as a guiding role for the fault diagnosis of system. A. Introduction With more and more widely used AMB in the industrial field, bigger challenge is emerging, such as the object supported of increasingly complex, both the size and the load capacity larger than before, and the speed of rotation higher, so many aspects have to be improved, especially for the control algorithm of the whole system. The normal amount of vibration displacement as the only parameter for the control system has been unable to meet the needs of the system. For some special industrial requirements, just like the 10MW high temperature gas-cooled test module reactor (HTR-10GT), designed by the INET of Tsinghua University, take the AMB as the support bearings, which is 3.5 meter long rotor, 230mm diameter, and 15000rpm rated speed, at the same time the system has to through the second bending critical. The result is that the large vibration would be happened during speeding up process for such a slender rotor, and even causes the rotor instability. In order to solve the problem and achieve to control precisely, for the HTR-10GT, the phase and speed of real-time need to be added for the AMB system. Furthermore, speed and phase also is the important index for the monitoring system and fault Diagnosis System[1]. Speed sensor can be divided into eddy current, magnetic, laser type, etc., which one you select according to the specific application environments. Although speed sensor technology is mature, but once applied in specific industrial area, especially in the reactor environment, the sensor must be high reliability. Due to the speed sensor fails in case, it will have an impact on the control algorithms, especially at high speed operating conditions, the consequences could be disastrous. The speed sensor is mainly composed of probes, cables and demodulation circuit three parts. Wherein the probe measures the state of the rotor, through long-term transmission cable, and demodulates by the hardware circuit, then the final specific speed and phase is calculated by the software. Taking into account the reliability issues, speed measurement system of AMB, is made of three sets of sensor in redundancy relationship. There are some kinds of ways to calculate the reliability of the redundant system[2-5], such as Neural network, Fault tree analysis and Markov process. Among these, Markov model is the intuitive and effective method of solving complex redundant system. So in this paper, Markov chain is introduced to solve complex redundant speed measurement redundant units system reliability. B. Markov Process Reliability refers to the capabilities of the whole system especially in the specified conditions in a specified time to complete required function, and the Markov process is a useful tool to calculate it. For the Markov process, some definition is given: Assuming ( ) { , } t X t T is the random process, and if for any positive integer n , 1 2 n t t t , 1 1 1 1 { ( ) , , ( ) } 0 n n P X t x X t x ,then