Development Of Intention Turning Bed Control System Based On Lying Posture Monitoring

With the intensification of population aging,more and more disabled elderly people are bedridden.Under the pressure of gravity,keeping the hips and back in the same lying position for a long time can easily cause pressure ulcers.In order to prevent the occurrence of pressure ulcers,the traditional method is that the nursing staff regularly use a hand-cranked or electric turning bed to roll over the disabled elderly,which not only increases the workload of the nursing staff,but also causes great harm to themselves due to passive rolling over.It is of great significance to realize the prevention of pressure ulcers on how to complete the self-controlled turning of the disabled group.Therefore,while designing the control system of the intention to turn over the bed,we also carried out in-depth research on the recognition of lying posture and the intention to turn over,and proposed a machine learning method based on the extreme learning algorithm to identify the lying posture,and a decision tree based on ID3 correlation BP neural network algorithm recognition.A deep learning approach to rollover intent.The main research contents and research results obtained are as follows:1)The software and hardware design of the intended turning bed control system is carried out.The human-computer interaction interface of the body pressure acquisition system was developed,and the intensity map of the back pressure and the two-dimensional pressure array were drawn.The back pressure intensity map can determine the lying posture of the disabled group,and the two-dimensional pressure array can determine the intention to turn over.The body pressure data is collected by using the flexible pressure sensor array and data acquisition card;in order to receive and transmit the signal,the turn-over information is uploaded through the Wi Fi module;the motor drive module is developed based on the STM32 controller.2)Research on machine learning method for recognizing lying posture based on extreme learning algorithm.A flexible pressure sensor array was used to obtain a static lying position pressure image,and 16 kinds of image eigenvalues were extracted,including 9 kinds of geometric eigenvalues,4 kinds of energy eigenvalues,and 3 kinds of color eigenvalues.By normalizing the eigenvalues Processing,the extreme learning algorithm was introduced to train the eigenvalues.The final correct rate of recumbent posture recognition reached 98.75%,realizing the recumbent posture recognition of disabled groups on the bed.3)Research on the method of identifying turning intention based on decision tree ID3 correlation BP neural network algorithm.The ID3 decision tree algorithm is proposed as the criterion for the information gain value to select the test attribute,and the maximum correlation of the turning intention is screened out through the information gain value.According to the two-dimensional pressure array,the pressure value extraction was completed when the experimenter turned left and right while lying down;the threshold value extraction of the turning intention by the BP neural network algorithm was introduced.It is obtained that the pressure array is collected with a delay of 1.2s in the population of different weights.When the difference between the left and right satisfies A-B>K,it is judged as turning left;when |A-B|?K is met,it is judged as lying flat;when A-B<-K is met,it is judged as Turn right(K is the critical threshold for turning over).4)Developed a long-distance wireless communication protocol using Huawei cloud server and MQTT communication.The MQTT wireless communication protocol replaces the conventional serial transmission and realizes the function of remote control of the turning bed.5)The experimental results were analyzed.It is concluded that the turning angle of the man is 45° when the turning time of the turning bed is 10?15s,the pause time after turning over is 15 s,and the average value of the delay response error is 0.21s;when the turning time of the woman is10 s,the turning angle is 45°,The pause time after turning over is 15 s.The conclusion that the average error of the entire delayed response for girls is 0.19 s.Through SD subjective evaluation,it was determined that the experimenter's physical and mental fatigue were significantly improved before and after turning over.It is concluded that the intentional rollover bed can effectively relieve the physical and mental fatigue of the inconvenient group when they lie down for a long time.