CSP-CNN model for traffic accident severity prediction and its modeling method

Abstract

The invention discloses a traffic accident severity prediction CSP-CNN model and its modeling method. The CSP-CNN model includes a model input layer, the model input layer inputs traffic accident dataconverted gray image set, and carries on convolution calculation to its input convolution layer, obtains the feature vector extracted from the last convolution layer, and inputs the feature vector tothe whole connection layer; a CNN model includes model input layer, model input layer inputs traffic accident data converted gray image set, and carries on convolution calculation to its input convolution layer, obtains the feature vector extracted from the last convolution layer, and inputs the feature vector to the whole connection layer. The full connection layer performs flatten operation onthe input eigenvectors, converts them into one-dimensional vectors and performs linear processing. The full connection layer contains three hidden units, and outputs three linear processing results tothe model output layer. Three traffic accident severity levels are set in the output layer of the model, and the Softmax activation function is used to predict the traffic accident severity. The invention fully considers the space-time relationship, the combination relationship and the deeper internal relationship between the traffic accident characteristics, and predicts the traffic accident severity.

Description

technical field [0001] The invention belongs to the technical field of data mining, and in particular relates to a traffic accident severity prediction model based on deep learning and a modeling method thereof. Background technique [0002] Every year, more than 1.25 million people around the world lose their lives as a result of road traffic accidents, and an additional 20 to 50 million suffer non-fatal injuries, many of them disabled. Road traffic injuries cost individuals, families, and entire countries economically, with road traffic collisions costing up to 3% of GDP in most countries. [0003] Accident severity prediction is one of the important steps in accident management, which provides important information for emergency personnel to assess the severity of accidents, evaluate the potential impact of accidents, and implement effective accident management procedures. Since the work of correctly predicting the severity of traffic accidents will provide extremely imp...

AI Technical Summary

Problems solved by technology

However, the above related works have not considered and explored the space, combination and deeper internal relationship among the features that affect the severity of traffic accident casualties in detail.

Images