Adaptability analysis method of geomagnetic map based on adaptive weighted PSO-BP model

In order to provide reliable support for AUV geomagnetic matching navigation and improve the accuracy of geomagnetic map adaptability analysis, this paper proposes a geomagnetic map adaptability analysis method based on the wPSO-BP model. By calculating and selecting the geomagnetic map characteristic parameters, through the main The component analysis method selects the calculated geomagnetic characteristic parameters, identifies and extracts the most representative principal components. Through the adaptive weighted particle swarm optimization-back propagation (wPSO-BP) model construction, the identified principal components are input into the constructed wPSO-BP model, and the geomagnetic map adaptability evaluation value is output, and then the map map adaptive area/non-adaptive area classification results are obtained. Realize effective dimensionality reduction of geomagnetic map characteristic parameters, and then achieve accurate analysis of geomagnetic map adaptability by constructing a wPSO-BP neural network model to achieve accurate classification of geomagnetic map adaptation areas/non-adaptation areas. The simulation results show that the wPSO-BP neural network model has higher prediction accuracy and shorter running time, which significantly improves the algorithm performance and becomes an effective means for evaluating the adaptability of geomagnetic maps.