Data processing: artificial intelligence – Adaptive system
Reexamination Certificate
2001-07-03
2004-06-08
Davis, George B. (Department: 2121)
Data processing: artificial intelligence
Adaptive system
C706S010000, C700S226000
Reexamination Certificate
active
06748370
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an integrated cellular network structure that can be programmed to solve partial derivative differential equations.
In particular, the invention relates to an integrated cellular network structure that can be programmed to solve partial derivative differential equations in order to control a phenomenon of diffusion or propagation of electric drive pulses for robot actuators.
2. Description of the Related Art
Reported in the pertinent literature are several examples of possible realizations of single-layer cellular neural networks primarily intended for the real-time processing of images.
By way of example, the report of the conference MicroNeuro '99, “CNNUC3: A Mixed-Signal 64×64 CNN Universal Chip” by G. Liñan, S. Espejo, R. Domínguez-Castro, E. Roca, Á. Rodriguez-Vasquez, is reported among such several examples.
Another noteworthy prior solution to perform a single-layer cellular neural network is disclosed in the U.S. Pat. No. 5,140,670 to L. O. Chua and L. Yang.
Although in many ways advantageous and substantially achieving their objectives, these prior solutions are limited in applicability and difficult to adapt to the processing of partial derivative differential equations of the second order.
Research work in electronics and in physics of semiconductors is aimed at developing CMOS technology integrated systems that can solve, in real time, a particular type of second-order partial derivative differential equations, known as reaction-diffusion equations.
The importance of such equations derive from the biological research, and especially from the study of nervous tissues, where many phenomena of wave-propagation have been detected which can be readily described with mathematical models of the reaction-diffusion type. Besides the biological field, there are numerous phenomena, such as chemical reactions, combustion, etc., that exhibit the same characteristics.
As described in many documents, particularly in European Patent Application No. 0 997 235 in the name of the same applicant, cellular neural networks have become excellent tools for solving such equations, though not in their most traditional known versions.
The underlying technical problem of the present invention is to provide an analog and digital cellular neural network integrated circuital structure with such structural and functional features as to allow partial derivative differential equations to be generated and controlled, and the network initial conditions and control parameters to be programmed.
BRIEF SUMMARY OF THE INVENTION
The disclosed embodiments of this invention provide a network wherein the values of the parameters of equations to be generated can be defined, and the form of the structure re-defined. Thus, the whole available array or just one or a few portions thereof can be used with different parameters and forms. Also, since the phenomena to be generated may have different time constants, the typical frequency of an individual cell can be varied within a broad range (1 MHz to 1 Hz).
Furthermore, the embodiments of the invention are very different from conventional cellular neural networks not only by its structure but also in the respect of its programmability.
The features and advantages of an integrated structure according to the invention will be apparent in the following description of an embodiment thereof, given by way of example and not of limitation with reference to the accompanying drawings.
REFERENCES:
patent: 5680070 (1997-10-01), Anderson et al.
patent: 5691664 (1997-11-01), Anderson et al.
Arena, P. et al., “Reaction-Diffusion CNN chip. Part1: IC Implementation,”IEEE International Symposium on Circuits and Systems, Geneva, Switzerland, May 28-31, 2000, pp.III-419—III-422.
Arena, P. et al., “Autowaves for Motion Control: A CNN Approach,”Proceedings of the 1998 IEEE International Symposium on Circuits and Systems, Monterey, CA, USA, May 31-Jun. 3, 1999, pp.III-163—III-166, 1998.
Arena Paolo
Branciforte Marco
Di Bernardo Giovanni
Occhipinti Luigi
Davis George B.
Jorgenson Lisa K.
Seed IP Law Group PLLC
STMicroelectronics S.r.l.
Tarleton E. Russell
LandOfFree
Integrated cellular network structure that is programmable... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Integrated cellular network structure that is programmable..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Integrated cellular network structure that is programmable... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3314747