A Multi-Axis Motion And Force Cooperative Control System For Dynamic Foot Biomechanical Simulator

The foot and ankle system in biomechanics is of key importance to theresearch of mechanism of ankle diseases, artificial prosthesis, orthopedic,development of shoes and related fields. It helps the study of mechanicsbetween different components of foot and ankle system including relativemotion and interaction force muscle force during a normal gait cycle andother complex movements. In order to implement the research, multifunctionfoot and ankle biomechanical measurement devices have been employed tofurther our understanding of the foot and ankle’s normal and pathologicfunction.A novel robotic foot and ankle biomechanical simulator (gait simulator)was developed to simulate both motion and force characteristics in a gaitcycle with5Degree-of-freedom (DOF). Major control challenges of the gaitsimulator includes multi-input multi-output, non-linear, coupling of force anddisplacement and multi-performance indicators. This paper has achieved tasksincluding modeling of the gait simulator, validation of the model, controlstrategy analysis, computer simulation, development of software andhardware and experiments of control algorithm.Firstly, based on the understanding of one gait cycle, the author analysedthe relationship between multi-input (the displacement curves of multi motors)and multi-output (ground reaction force in three dimension and muscleforces). A simplified kinetic model of the gait simulator was built up inMatlab and sensibility analysis of some minor components was completed. By comparing the output between model and practical gait simulator underthe same displacement input, the model was verified and simplified.Secondly, based on the model above which is multi-input multi-output,strong coupling, PID type of iterative learning control (ILC) and a fuzzy logicparameter tuning controller was proposed to control the simulation system toreach the technical specifications. Computer simulation of the control strategywas completed in Matlab.Thirdly, hardware and software system of the gait simulator wereimplemented. Hardware including motors, drivers, motion control PCI board,data collection PCI board and sensors was selected according to the systemrequirements. Software was developed on MFC platform with functions offorce data collection, control of motion and force application, analysis ofexperiments and a visual interface.Finally, ILC experiments with and without fuzzy tuning logic were doneon the gait simulator. An evaluation and a comparison were made on theproposed control strategy. The control strategy was validated in the simulationplatform with a prosthetic foot. The result reviewed that our novel gaitsimulator with high verified repeatability could complete the motion andforce loading process in5seconds with less than550N body weight in arelatively high similarity of human behavior. The tracking curves ofF_z andF_y can converge to the target ones within4%and20%RMS error using ILCrespectively.The proposed control strategy greatly improved intelligence of the gaitsimulator and provided a good foundation to further improve the simulationspeed and accuracy. This work was supported by the Natural ScienceFoundation of China through the research project (No.81071234). It will givea reference to related research of multi-input multi-output biomechanicalsimulator.