Search for: All records

As neural networks have become increasingly prolific solutions to modern problems in science and engineering, there has been a congruent rise in the popularity of the numerical machine learning techniques used to design them. While numerical methods are highly generalizable, they also tend to produce unintuitive networks with inscrutable behavior. One solution to the problem of network interpretability is to use analytical design techniques, but these methods are relatively underdeveloped compared to their numerical alternatives. To increase the utilization of analytical techniques and eventually facilitate the symbiotic integration of both design strategies, it is necessary to improve the efficacy of analytical methods on fundamental function approximation tasks that can be used to perform more complex operations. Toward this end, this manuscript extends the design constraints of the addition and subtraction subnetworks of the functional subnetwork approach (FSA) to arbitrarily many inputs, and then derives new constraints for an alternative neural encoding/decoding scheme. This encoding/decoding scheme involves storing information in the activation ratio of a subnetwork’s neurons, rather than directly in their membrane voltages. We show that our new “relative” encoding/decoding scheme has both qualitative and quantitative advantages compared to the existing “absolute” encoding/decoding scheme, including helping to mitigate saturation and improving approximation accuracy. Our relative encoding scheme will be extended to other functional subnetworks in future work to assess its advantages on more complex operations. 

This study introduces a novel neuromechanical model of rat hindlimbs with biarticular muscles producing walking movements without ground contact. The design of the control network is informed by the findings from our previous investigations into two-layer central pattern generators (CPGs). Specifically, we examined one plausible synthetic nervous system (SNS) designed to actuate 3 biarticular muscles, including the Biceps femoris posterior (BFP) and Rectus femoris (RF), both of which provide torque about the hip and knee joints. We conducted multiple perturbation tests on the simulation model to investigate the contribution of these two biarticular muscles in stabilizing perturbed hindlimb walking movements. We tested the BFP and RF muscles under three conditions: active, only passive tension, and fully disabled. Our results show that when these two biarticular muscles were active, they not only reduced the impact of external torques, but also facilitated rapid coordination of motion phases. As a result, the hindlimb model with biarticular muscles demonstrated faster recovery compared to our previous monoarticular muscle model.