Magnetometer Calibration Method Based on Genetic Algorithm Optimization and Improved BP Neural Network

Abstract

The invention discloses a magnetometer correcting method with an optimized and modified BP neural network based on a genetic algorithm. An error model of a magnetometer is converted into the three-layer modified BP neural network (namely the parameters of the error model of the magnetometer serve as weights of the BP neural network for optimized estimation), then the weights of the modified BP neural network are optimized in real time through the genetic algorithm, and therefore correction and compensation of the magnetometer are achieved; the structural principle of the modified BP neural network is shown in the picture 2. The correction method can effectively improve the accuracy of the magnetometer, further achieves high-accuracy positioning and navigation, lowers the processing complexity of a computer, obtains a good effect, has the advantages of being high in practicability, real-time performance and reliability, low in cost, high in accuracy, simple in calculation process, high in stability and the like, and can effectively improve the accuracy of the course angle of a navigation system.

Description

technical field [0001] The invention relates to a method for correcting a magnetometer based on a genetic algorithm to optimize and improve a BP neural network, which can correct the magnetometer in a navigation system. Background technique [0002] In the application of the navigation system, whether the navigation system can provide the accurate attitude and heading angle information of the carrier is very important. For an inertial measurement unit (IMU) composed of only accelerometers and gyroscopes, the accuracy of the attitude and heading parameters output by the IMU cannot meet the requirements of some navigation systems due to the errors and drift of the accelerometers and gyroscopes themselves. The attitude and heading measurement system composed of inertial measurement unit (IMU) and magnetometer can improve the accuracy of attitude angle and heading angle. Magnetometer technology is widely used abroad, but there are still some problems in magnetic field compensat...

AI Technical Summary

Problems solved by technology

However, most methods need to transform the magnetometer error model into an ellipsoid model, and then use some methods to fit the ellipsoid to obtain the parameters of the ellipsoid model. It reduces the processing complexity of the computer, and at the same time, it needs to add restrictions when fitting the ellipsoid

There are also some methods that can only perform static calibration and compensation on the magnetometer, which do not have good real-time performance

Images