Design of an Artificial Neural Network Algorithm for a Low-Cost Insole Sensor to Estimate the Ground Reaction Force (GRF) and Calibrate the Center of Pressure (CoP)

Abstract

As an alternative to high-cost shoe insole pressure sensors that measure the insole pressure distribution and calculate the center of pressure (CoP), researchers developed a foot sensor with FSR sensors on the bottom of the insole. However, the calculations for the center of pressure and ground reaction force (GRF) were not sufficiently accurate because of the fundamental limitations, fixed coordinates and narrow sensing areas, which cannot cover the whole insole. To address these issues, in this paper, we describe an algorithm of virtual forces and corresponding coordinates with an artificial neural network (ANN) for low-cost flexible insole pressure measurement sensors. The proposed algorithm estimates the magnitude of the GRF and the location of the foot plantar CoP. To compose the algorithm, we divided the insole area into six areas and created six virtual forces and the corresponding coordinates. We used the ANN algorithm with the input of magnitudes of FSR sensors, 1st and 2nd derivatives of them to estimate the virtual forces and coordinates. Eight healthy males were selected for data acquisition. They performed an experiment composed of the following motions: standing with weight shifting, walking with 1 km/h and 2 km/h, squatting and getting up from a sitting position to a standing position. The ANN for estimating virtual forces and corresponding coordinates was fitted according to those data, converted to c script, and downloaded to a microcontroller for validation experiments in real time. The results showed an average RMSE the whole experiment of 31.154 N for GRF estimation and 8.07 mm for CoP calibration. The correlation coefficients of the algorithm were 0.94 for GRF, 0.92 and 0.76 for the X and Y coordinate respectively.

Keywords: artificial neural network (ANN); center of pressure (CoP); force sensing resistor (FSR); ground reaction force (GRF).

Figure 1

Design of the low-cost flexible insole pressure measurement sensor with six FSR sensors. Gray circles with wires show the FSR sensors. White bolded numbers indicate the anatomically divided areas. Blue squares mean the areas of the virtual forces that contain the FSR sensors in them.

Figure 2

Flow chart of the ANN algorithm for estimating the GRF and calibrating the CoP of the insole.

Figure 3

Picture of experimental setup. All experiments were done wearing the F-Scan system in-shoe sensor and low-cost FSR foot sensor simultaneously. (1), (2), (3) represents the shoe, developed FSR foot sensor, and the F-Scan in-shoe sensor respectively. And they are piled up in order.

Figure 4

Results of Algorithm Fitting with Different Layers of Hidden Neurons for Virtual Forces. Left graph presents the mean of error resulted by each number of hidden layers and right graph is for correlation coefficient. Each colored circle means each virtual force (Blue-$F_{ET}$, Orange-$F_{ER}$, Grey-$F_{EL}$, Yellow-$F_{EH}$, Purple-$F_{EC}$, Green-$F_{ELT}$ ).

Figure 5

Results of algorithm fitting with different numbers of data frames. Yellow circle means the RMS error and purple one means the time duration of each layer.

Figure 6

(a) The CoP trajectory plot during the scan experiment by the WMA, VFA and F-Scan. The gray circles with bold edges indicate the locations of the FSR sensors. (b) One Chosen GRF trajectory plot during the scan experiment by the VFA and F-Scan.

Figure 7

(a) One chosen CoP trajectory plot during the stance phase of the 1 km/h gait cycle experiment by the WMA, VFA and F-Scan. (b) One chosen GRF trajectory plot during the stance phase of the 1 km/h gait cycle experiment by the VFA and F-Scan.

Figure 8

Bar graph of the average RMSE during experiments of all subjects by the VFA (Yellow bar) and the WMA (Green bar).

Figure 9

One chosen frame of CoP trajectory plot during the scan experiment by WMA (No Algorithm) with orange point, the VFA (Algorithm) with green point, and the reference data (F-Scan) with pink point.