Application Of Voltage Sag Detection Algorithm In Self-starting System Of Low-voltage Motor Group

Voltage sag is a very serious power quality problem for petrochemical enterprises,mainly reflected in the key low-voltage motor shutdown due to voltage sag,once there is no immediate start-up,It will cause great economic loss to the enterprise because of the interlock stop.The self-starting system of low-voltage motor group shakes is widely used in petro-chemical enterprises because it can restart the out-of-operation motor in time after the voltage sags occur.However,the low-voltage motor group start-up system of Ningxia Petrochemical Company is poor in practicability,mainly reflected in the failure to capture the voltage sags within 100ms.In order to change this situation,it is necessary to analyze the reasons for its poor practicability and upgrade the system on this basis.Therefore,this paper firstly analyzes the hardware and software algorithms of the system,and analyzes the working characteristics of the main hardware and the principle of the algo-rithm used in the system.It is concluded that the main reasons for the poor practicability are the selection.of system hardware and the improper selection of voltage sag detection algo-rithm.Secondly,the characteristics of voltage sags caused by three common reasons,such as short circuit faults,are studied by Matlab simulation,and the starting and stopping time of sags,the depth of voltage sags are discussed.The principle of several traditional detection methods is analyzed and evaluated by three characteristic variables of phase jump.In view of the working principle of the self-starting system for low-voltage motors,this paper focuses on the advantages and disadvantages of the traditional algorithm in detecting the starting and stopping moments and the depth of temporary drop.In view of the problem of poor real-time performance in traditional voltage sag detection through the detection of sag depth,a detection method of starting and stopping time based on edge detection operator of mathematical morphology image is proposed.The method is com-pletely independent of the sag depth,and the detection error of the starting and stopping time of voltage sag is verified to be 0.2 m,s by simulation.In view of the fact that the single-phase dq transform algorithm needs 1/4 cycle delay and too few sampling points of voltage wave-form,it is more sensitive to noise and harmonics when detecting the transient depth.A volt-age sag depth detection algorithm based on improved single-phase dq transform is presented in this paper.The algorithm can arbitrarily reduce the delay angle of constructing virtual three-phase system and realize the rapid detection of voltage sag depth.It is verified by sim-ulation that the detection delay of voltage sag depth is less than 2 Ms,and the problem of low noise and harmonic suppression ability is avoided by adjusting the delay angle to reasonable value.Based on the in-depth study of voltage sag detection algorithm and hardware reasons,a voltage sag detection module based on ARM is proposed.The two algorithms are used to de-tect the starting and stopping time of voltage sag and the depth of sag,respectively.At the same time,based on this module,the self-starting system of low-voltage motor group shaking in Ningxia Petrochemical Company is reformed.The field operation shows that the modified system runs smoothly and can capture voltage sags within 10ms.