Constrained chaos in three-module neural network enables to execute multiple tasks simultaneously

Area of Science:

• Computational neuroscience
• Robotics
• Chaos theory

Background:

• Neural networks with feedforward inter-module structure can potentially perform multiple tasks.
• Constrained chaos can be introduced into neural networks to explore novel functionalities.

Purpose of the Study:

• To investigate the potential of constrained chaos in a three-module neural network for simultaneous multiple task execution.
• To apply this concept to a robot actuator control system.

Main Methods:

• A three-module neural network (sensory, inter, motor) with binary neurons was designed.
• Chaos was introduced by reducing intra- and inter-module connectivity.
• Computer experiments were used to simulate a robot actuator controlled by the network.

Main Results:

• The system generated chaotic motion signals in the motor module.
• Simultaneous application of two distinct inputs resulted in the simultaneous execution of two corresponding tasks.
• Chaotic activities persisted even under strong dual-input conditions.

Conclusions:

• Constrained chaos in redundant neural systems facilitates simultaneous multiple task execution, mimicking brain capabilities.
• This approach offers a novel method for developing sophisticated robot control systems.