Truck load capacity image recognition model generation method and truck load capacity recognition method

Abstract

The invention belongs to the technical field of vehicle load capacity measurement methods, and particularly relates to a vehicle load prediction method based on an image LSTM and Kalman model. The method comprises the following steps: collecting picture data of a plurality of same trucks on site in the whole loading process; converting a picture format, carrying out noise reduction processing on the image, and dividing the processed data into a training set and a test set; constructing the LSTM and Kalman model, inputting the training set data into the LSTM and Kalman model for training, and forming a training model to predict the load capacity of the vehicle; inputting test set data into the trained LSTM and Kalman network structure model for verification, and calculating prediction accuracy; and optimizing and adjusting parameters of the LSTM and Kalman network structure model through the test set, improving the detection precision, and using the finally adjusted network model as a final large truck load capacity image recognition model to realize prediction of the vehicle load capacity.

Description

technical field [0001] The invention belongs to the technical field of vehicle load measurement methods, in particular to a vehicle load prediction method based on an image LSTM and a Kalman model. Background technique [0002] How to accurately and efficiently detect the load data of large trucks is crucial to improving traffic safety and realizing large-scale truck load monitoring, and is of great significance to the vigorous development of road transportation construction and economy. Usually, the load detection of trucks uses strain gauges as a detection device installed on the suspension of the truck, and the load is detected by detecting the elastic deformation of the steel plate. However, as a detection device, strain gauges need to output data by detecting deformation. It is necessary to change the suspension structure of the truck to make it fit closely with the steel plate of the truck. Therefore, long-term maintenance is required to scrape off the anti-rust paint ...

AI Technical Summary

Problems solved by technology

However, the BPNN method has poor timeliness for real-time prediction of images, and cannot solve the problems of data delay and data error.

Images