A Support Vector Machine-Based Correction Method for Measuring Voltage Value of Insulator Inspection Robot

Abstract

The invention relates to a method for correcting the voltage value measured by an insulator detection robot based on a support vector machine. According to the distribution parameter model of the insulator and the equivalent path model of the insulator detection robot, the situations of different low and zero value insulators are simulated, and the insulator strings are calculated respectively when there is no robot. The voltage distribution value and the insulator string voltage distribution value measured by the robot on each piece of insulator, the two sets of data are used as samples to form a sample library; select the sample set, use the support vector machine model for training, correct the actual voltage distribution, and optimize the model parameters Complete the model training; use the trained model to correct the measurement data, and the real voltage distribution of the insulator string can be obtained. This method can obtain the complex nonlinear relationship between the measured voltage value and the actual insulator string distribution voltage through training and learning of sample data under typical defect conditions, and accurately correct the measured voltage values ​​under normal conditions and different low and zero value insulators .

Description

technical field [0001] The invention relates to the technical field of insulator robot detection for transmission line insulator strings, in particular to a support vector machine-based insulator detection robot measurement voltage value correction method. Background technique [0002] Transmission line insulators work in an outdoor environment and are affected by harsh weather or industrial pollution for a long time. They may be eroded by dust, natural hydrochloric acid, smog, acid rain, industrial waste gas and other pollutants. As the operation time prolongs, the insulation performance will decrease. In order to ensure the safe operation of transmission line insulators, it is necessary to regularly detect the deterioration of porcelain insulators. The traditional detection method is mainly manual detection, using the spark fork method, voltage distribution method, etc. to identify insulators. It is very inconvenient for long strings of insulators due to long operating rod...

AI Technical Summary

Problems solved by technology

[0003] However, the insulator detection robot may change the electric field distribution of the original insulator string during the measurement process

The equivalent electrical model of the detection robot is composed of the capacitance between the probes, the measurement resistance and the equivalent impedance of the measurement circuit. When the detection robot measures a piece of insulator, it is equivalent to connecting a branch in parallel next to the insulator, so that the overall impedance is reduced, thus The partial voltage decreases, the measured voltage value is smaller than the real value, and there is a certain deviation, which may cause misjudgment of low and zero value insulators

Moreover, when the insulator robot detects the voltage of each piece of insulator in the insulator string, the position is different, so the equivalent electrical network is constantly changing, so the measured distribution voltage value of the entire string and the distribution voltage value of the actual insulator string are complicated and nonlinear relation

Images