Fall detection system and method based on data fusion and BP neural network

Abstract

The invention provides fall detection system and method based on data fusion and BP neural network. Being set up inside a shoe, the fall detection system based on the data fusion and the BP neural network is used for predicting fall. The fall detection system based on the data fusion and the BP neural network comprises a first subsystem and a second subsystem which are separately arranged in a shoe for a left foot and a shoe for a right foot; each of the first subsystem and the second subsystem comprises four membrane pressure sensors which are arranged inside an insole and used for collectingplantar pressure data, an accelerometer gyroscope module which is used for collecting foot acceleration data and foot angle velocity data, a WiFi module which is used for carrying out data transmission, a main control chip which is used for carrying out data processing and a battery module which is used for supplying power, wherein the four membrane pressure sensors are respectively arranged on the insole at positions corresponding to the first toe, the first metatarsal bone, the fourth metatarsal bone and the heel; and the accelerometer gyroscope modules, the WiFi modules, the main control chips and the battery modules are all arranged outside shoe uppers or integrated inside shoe sole. The invention further provides a detection method of the fall detection system based on the data fusion and the BP neural network for predicting fall.

Description

technical field [0001] The invention relates to a fall detection system, in particular to a fall detection system and method based on data fusion and BP neural network. Background technique [0002] Falls are an important cause of hip fractures in the elderly. With the increase of the aging population in our country, people pay more and more attention to the fall problem of the elderly. The uncertainty of its occurrence and the serious consequences have made a real-time monitoring fall prevention device a current research topic. hotspot. The existing fall detection system methods are generally divided into three types: machine vision-based fall detection, environmental sensor-based fall detection, and wearable fall detection systems. The first two have great limitations because they can only be performed in an indoor environment. The latter is easy to wear, low in cost, and very suitable for productization, but the existing wearable systems are generally based on the thre...

AI Technical Summary

Problems solved by technology

The first two have great limitations because they can only be carried out in an indoor environment.

The latter is easy to wear, low in cost, and very suitable for productization, but the existing wearable systems are generally based on the threshold method, which has a simple structure, but poor sensitivity and specificity, and the threshold is usually based on Specific testers set the maximum or minimum value, which has poor universality. At the same time, most of the existing inventions alarm after the user falls, and cannot respond before the fall. Research shows that the duration of the fall is 300ms-500ms , when the fall has occurred or the user has been injured and then the alarm is issued, the safety of the user cannot be guaranteed

Images