Research on Normal Human Plantar Pressure Test

FSR400 pressure sensor, nRF905 wireless transceiver and MSP40 SCM are used to design the insole pressure collection system, LabVIEW is used to make HMI of data acquisition, collecting a certain amount of normal human foot pressure data, statistical analysis of pressure distribution relations about five stages of swing phase during walking, using the grid closeness degree to identify plantar pressure distribution pattern recognition, and the algorithm simulation, experimental results demonstrated this method feasible.


Introduction
When the body in accordance with normal gait, plantar only directly contact with the ground, gait would have plantar pressure.The various factors of human physiology, diseases, etc. will affect the person's gait, it studies the body's normal plantar pressure when walking, that have important reference value for researching gait analysis, medical rehabilitation, clinical and smart shoes, and so on [1] .
Based on the physical structure and division of the human foot plantar anatomical region, In accordance with the left and right foot insoles heel, arch, first metatarsal, the second metatarsal, metatarsal 3rd, 4th metatarsal, 5th metatarsal and the first toe area, each placed a film pressure sensor FSR400, each foot collects eight points force [2] .Distribution of plantar pressure sensor shown in Figure 1  First, according to the design requirements, the serial port was initialized, set the serial port was set and the baud rate was set to 9600b/s, 8 data bits, 1 stop bit, no parity.Second, the control system was set to start/pause/resume/stop/sampling interval frequency; Setting pressure threshold, when the pressure is greater than this value, the corresponding channel curve becomes green, setting the display area Y-axis display range; Setting data save path, note that the file must be in ".txt" as suffix.
PC sends a start command to the microprocessor "Start", After the microprocessor receives, sends a signal "DSR" of "ready to send" to the PC, and wait for the PC feedback signal "R" of "ready to receive", and then start to send a frame of data, the transmission is completed,  shown in Table 1.The process of between twice land of the same foot heel is a gait cycle.A gait cycle can be divided into the swing phase and the stance phase.When the swing phase, Mid Stance, (4) Heel Rise, (5) Toe Contac t [3][4] .
Due to the experimental subject's weight, height and other differences, the sensor data for the individual differences in subjects is different, so if only thinking of the specific sensor measurements, it is difficult to judge the subjects in which stance phase, so using fuzzy mathematical to process the experimental data, and using the membership function to represent the relative size of the measured sensor data.For each measurement sensor, the corresponding membership functions such as formula (1): x is the sensor values subject standing still, a is as the adjustment factor, the value of a is the greater, the membership function P corresponding to the curve is steeper.

Recognition
Fuzzy Recognition is often used in two ways: one is the maximum membership degree principle, also called direct method, which is mainly used to identify a single target; The other is closeness principle, also known as the indirect method, which is generally used for the cluster target identification.Closeness is the proximity of two fuzzy sets of metrics.In this paper, the principle of elective near close degree of grid to identify the subjects in which the stance phase (distribution mode).

Simulation and Analysis
In order to verify the effectiveness of this method, we have compiled a handler in Matlab platform, added 15 new subjects, collected in plantar pressure data when they were five kinds of stance phase, substituted the data to the program and got automatic recognition results as shown in Table 2.

Figure 2 .
Figure 2. Insole plantar pressure measurement system hardware block diagram Wireless transceiver module nRF905 built a complete communication protocol and CRC check circuit, and from on-board hardware automatically Manchester encoding / decoding, Just to complete all of the radio transmission through the SPI interface.Use of wireless communications technology, the collected signal was to the initial point , waiting for the next start command.After the PC receives the data, first by "String to Byte Array" command to convert the string into an array, and by measuring the file command data to the "pressure data .txt"file is saved in the appropriate folder while displaying real-time data 8 channels, each channel can display the current value of the measured data.Data Acquisition HMI shown in Figure 3. 02060-p.2

Figure 3 .
Figure 3. Data Acquisition HMI leg off the ground and in the air, swinging forward.The stance phase with the lower limbs in contact with the ground to withstand the ground reaction force and a role in stabilizing the body.In the stance phase, can be divided into five stages (five kinds of distribution patterns):( 1) Initial Contact, (2) Loading Response,(3)

.
set of theory domain, constitutes a standard model library, it is to be a recognized fuzzy subset of domain U.If there the stance phase are corresponding to five pressure distribution patterns.According to these five kinds of distribution patterns, the plantar pressure data of the 67 cases subjects are divided into five categories, fuzzy them and getting five sample sets, each element of these sample sets is 8-dimensional, is corresponding to fuzzy membership of 8 sensors.Each sample set is statistical and averaging, During walking, 8 sensor is placed on the foot to collect of the plantar pressure data, these data can be converted into membership, characterizing pressure value relative size.Let sample, i b is a element of set, each element consists of eight data, to the ith element there has

From
the experiment, an average recognition rate was 82.6%, we can see the method on human plantar pressure distribution pattern has better recognition performance.Because plantar pressure distribution pattern and the calculation of five standard distribution models ^1 2 3 4 5 , , , , A A A A A are on the basis of statistical data, sample number is insufficient that affects to increase the recognition rate.It is necessary to improve sample 02060-p.4number and increase the sample diversity in the follow-up study.

Insole Plantar Pressure Measurement System 2.1 Hardware Design of Insole Plantar Pressure Measurement System
.
Figure 1.The distribution of plantar pressure sensor 2 thereby to obtain pressure information, and the greater the pressure, the lower the resistance.The voltage signals of 8 sensors were what we wanted to measure.Because the human body 98% of plantar pressure signal DOI: 10.1051/ C Owned by the authors, published by EDP Sciences, 201

Table 1 .
The normal district average

Table 2 .
The algorithm identifies results