Neuro-mechanical networks as an architecture for system design (original) (raw)

Design and configuration of neuro mechanical networks

Structural and Multidisciplinary Optimization, 2008

Neuro Mechanical Network (NMN) is a new concept of adaptronic character. The governing idea is to include geometry, topology, load carrying, energy transfer, actuating, sensing and control of a machine in one single mathematical state model, and thereby enabling a formulation of the design and configuration problem as an optimization problem. We have focused our attention on a type of NMNs consisting of what we call active trusses. For these we have established a state model and given a design optimization problem from which we have obtained numerical solutions. These solutions show that the approach has the possibility to suggest new families of designs that are superior to those of classical passive trusses. We also indicate how activation may result in singularities, the treatment of which is so far essentially an open problem.

Simulation of self organizing structures using neuro mechanical networks

The neuro mechanical network consists of a large number of one-dimensional elements connected into a topological graph of intelligent actuators in 2 or 3 dimensions. This forms a self actuating mechanical network that can be trained to perform certain tasks. In the analysis and training of such networks the time domain simulation of the network performance becomes important. Even though the basic components hardly exist in hardware at present, the study of such networks gives us interesting models to design and analysis the mechanisms of the near future using current technologies and engineering tools. The neuro mechanical network has a meaning also at a micro or even macro level in order to realize highly robust flexible actuator systems. Another potential use is for design of more conventional system, requiring a minimum of components. Furthermore it can be used as an explanatory model for some of the mechanics found in very complex biological systems, e.g. heart muscles.

Neuro-Mechanical Networks -Self-Organising Multifunctional Systems

Network systems have been subject to increasing interest during the recent decades. One of the more striking features of networks is that complex properties can emerge although the basic elements are very simple. This is particularly true in neural networks that has become a firmly established discipline for signal processing and control systems. There are also cellular automata, which is another branch where a great deal of interest is devoted.

Flexible body control using neural networks

Progress is reported on the control of Control Structures Interaction suitcase demonstrator (a flexible structure) using neural networks and fuzzy logic. It is concluded that while control by neural nets alone (i.e., allowing the net to design a controller with no human intervention) has yielded less than optimal results, the neural net trained to emulate the existing fuzzy logic controller does produce acceptible system responses for the initial conditions examined. Also, a neural net was found to be very successful in performing the emulation step necessary for the anticipatory fuzzy controller for the CSI suitcase demonstrator. The fuzzy neural hybrid, which exhibits good robustness and noise rejection properties, shows promise as a controller for practical flexible systems, and should be further evaluated.

A novel biomimetic actuator system

Robotics and Autonomous Systems, 1998

The design of a biomimetic actuation system which independently modulates position and net stiffness is presented. The system is obtained by arttagonisticaUy pairing contractile devices capable of modulating their rate of geometric deformation relative to the rate of deformation of a passive elastic storage element in series with the device's input source. A mechanical model is developed and the properties of the device are investigated. The theoretical results developed are then compared with experimental evidence obtained from a simple prototype model of the system. upon similar experimental results, Alexander [2] asserts that humans store energy in their Achilles tendons and the ligaments that support the arch of the foot, and that compliant legs and feet reduce the peak forces that occur when the foot strikes the ground. In a similar vein, Cavagna et al. suggest that running is essentially bouncing.

A Novel Neural Network Structure for Motion Control in Joints

—This paper proposes a new type of artificial neural network useful for motion control of end-point of a joint typically seen in biological systems. The network is based on a hybrid concept combining Adaptive Resonance Theory (ART) and Self Organizing Maps (SOMs). Basic concepts related to the new architecture and necessary algorithm to implement the network are presented in the paper. General applicability of the proposed method using two different kinds of joints for two different types of gradient functions has been presented. The algorithms have been implemented using R simulation language. Results of the implementation are also presented in the paper.

A biological neural network drives a robotic actuator

2006

During the past three years our group experimented the growth of networks of human neural stem cells on a MEA (Microelectrode Array) support. The neurons were stimulated by digital patterns and the output signals were analysed. In previous experiments, the neurons replied selectively to different patterns and showed similar reactions in front of the presentation of identical or similar patterns. Analyses performed with a novel Artificial Neural Network called ITSOM showed the possibility to decode the neural responses to different patterns. In the described experiment, the neurons are connected to a robotic actuator: simulated perceptions stimulate the neurons, that react with organized electric signals. The signals are decoded by the Artificial Neural Network, that drives a minirobot.

Control of Bionic Prosthesis using Neural Networks

Control of Bionic Prosthesis using Neural Networks, 2022

The integration of Artificial Intelligence and Machine Learning into the bionic prosthesis industry has caused subjects with mobility restrictions and inability to perform the simplest household tasks to see this technological advance with good eyes. Several anatomical and biomechanical functions once represented by the human body are now simulated due to rehabilitation aids having neural networks as their main suspect. The application of Artificial Intelligence technology has a huge impact on obtaining independent mobility and improves the quality of life of millions of amputees worldwide.

From Muscle to Brain: Modelling and Control of Functional Materials and Living Systems

This paper describes modelling and control of typical open systems: one is electroactive polymer gel and another is spinal nervous system. It is very im- portant to estimate the model based on their mechanisms in order to navigate the subjects into objective states. Firstly, the wave-shape pattern control method was proposed based on the gel model. Wave-shaped gels with varying curvature were obtained by switching the polarity of a spatially uniform electric field. Secondly, time series of images which represent distribution of somatic information inside the spinal cord were successfully obtained through measurement and computation utilizing somatotopic organization model of the spinal cord. The general problem underlying these studies is the degrees-of-freedom problem. Making use of the na- ture of functional materials or living systems through modelling their mechanisms helped us to solve the problem.

Actuator design using biomimicry methods and a pneumatic muscle system

Control Engineering Practice, 2006

An empirical and theoretical study is conducted on a special actuator termed ''pneumatic muscle'' (PM) being used in a force control system framework. Such an actuator has similarities to biological systems and has many advantages (extremely high power/ weight, power/volume and power/energy ratios). However, due to its inherent nonlinearities, this actuator suffers from poor position and force control. The study described here accomplishes three main goals. (1) A force control system is developed within an open and closed loop framework to emulate how biological systems work in an agonist-antagonist framework. (2) The PM used in the study has such strength that it excites the frame dynamics. This undesired dynamic response is then effectively cancelled using an impedance model control scheme. (3) The PM is demonstrated to both change length yet still produce force in a controlled manner.