Randomized Improved Correlation Matrix Memory in ANN-based Semi-Autonomous Robotic System

J.L.K. Soon and T.T.P. Lian (Malaysia)


Robotics, Artificial Neural Networks, Embedded System, Randomized Improved Correlation Matrix Memory


This paper presents the development of a semi-autonomous robotic system that operates based on Artificial Neural Networks (ANN). The ANN is a computational technique which models the way biological neurons work. ANN is used in pattern recognition, adaptive filtering, control sys tem and others. Application of embedded systems technol ogy such as microcontroller or programmable logic device (PLD) is necessary to control the system. Through wire less communication between the embedded system and a general-purpose computer, the system can be further em powered. With regard to the principle of equivalence of hardware and software, ANN can be developed in software or hardware form. The system is a prototype that serves as a foundation for further development for many appli cations including space exploration and services. An in novative improved version of Correlation Matrix Memory (CMM) learning algorithm capable of computing random optimized memory matrix and solving linearly inseparable exclusive or (XOR) logical operations was discovered and implemented. It is termed as Randomized Improved Cor relation Marix Memory (RICMM).

Important Links:

Go Back