A COMMERCIAL OFF-THE-SHELF IMPLEMENTATION OF AN ANALOG NEURAL COMPUTER

2008 ◽  
Vol 17 (02) ◽  
pp. 241-258 ◽  
Author(s):  
SANJAY K. BODDHU ◽  
JOHN C. GALLAGHER ◽  
SARANYAN A. VIGRAHAM
Keyword(s):  
Differential Equation ◽  
Neural Networks ◽  
Analog Circuits ◽  
Integrated Circuit ◽  
Electrical Circuit ◽  
Embedded Control ◽  
Accuracy And Precision ◽  
Application Specific Integrated Circuit ◽  
Commercial Off The Shelf ◽  
Digital Simulations

For most applications, analog electrical circuit implementations of continuous-valued neural networks have been abandoned in favor of digital simulations. This is not surprising, as both precision and accuracy can be more easily ensured in digital computers. Still, because they use far fewer transistors and support components, analog circuits can still be orders of magnitude smaller than their digital simulations. In some application, like micro-robotics and embedded control, one might be willing to tolerate less accuracy and precision for the size and power benefits. One would not under any condition, however, tolerate significant behavioral mismatches between the differential equation and electrical circuit forms of the neural networks in question. In this paper, we will present a design for an analog neural computer that embodies the commonly used continuous time recurrent neural network. We will show that the computer possesses excellent behavioral congruence to the differential equation form even in the presence of significant practical compromises. We will also discuss the implications of this work for both practical Commercial, Off-The-Shelf (COTS) and Application-Specific Integrated Circuit (ASIC) devices.

scholarly journals One-chip analog circuits for a new type of plasma wave receiver on board space missions

2017 ◽  
Vol 6 (1) ◽  
pp. 159-167 ◽  
Author(s):  
Takahiro Zushi ◽  
Hirotsugu Kojima ◽  
Hiroshi Yamakawa
Keyword(s):  
Plasma Wave ◽  
Analog Circuits ◽  
Integrated Circuit ◽  
Dynamic Range ◽  
Test System ◽  
Analog To Digital ◽  
Application Specific Integrated Circuit ◽  
New Type ◽  
Narrow Dynamic Range ◽  
Asic Chip

Abstract. Plasma waves are important observational targets for scientific missions investigating space plasma phenomena. Conventional fast Fourier transform (FFT)-based spectrum plasma wave receivers have the disadvantages of a large size and a narrow dynamic range. This paper proposes a new type of FFT-based spectrum plasma wave receiver that overcomes the disadvantages of conventional receivers. The receiver measures and calculates the whole spectrum by dividing the observation frequency range into three bands: bands 1, 2, and 3, which span 1 Hz to 1 kHz, 1 to 10 kHz, and 10 to 100 kHz, respectively. To reduce the size of the receiver, its analog section was realized using application-specific integrated circuit (ASIC) technology, and an ASIC chip was successfully developed. The dimensions of the analog circuits were 4.21 mm  ×  1.16 mm. To confirm the performance of the ASIC, a test system for the receiver was developed using the ASIC, an analog-to-digital converter, and a personal computer. The frequency resolutions for bands 1, 2, and 3 were 3.2, 32, and 320 Hz, respectively, and the average time resolution was 384 ms. These frequency and time resolutions are superior to those of conventional FFT-based receivers.

scholarly journals One-chip analogue circuits for a new type of plasma wave receiver onboard space missions

2016 ◽  
Author(s):  
Takahiro Zushi ◽  
Hirotsugu Kojima ◽  
Hiroshi Yamakawa
Keyword(s):  
Plasma Wave ◽  
Analog Circuits ◽  
Integrated Circuit ◽  
Dynamic Range ◽  
Test System ◽  
Analog To Digital ◽  
Analogue Circuits ◽  
Application Specific Integrated Circuit ◽  
New Type ◽  
Asic Chip

Abstract. Plasma waves are important observational targets for scientific missions investigating space plasma phenomena. Conventional fast Fourier transform (FFT)-based spectrum plasma wave receivers have the disadvantages of a large size and a narrow dynamic range. This paper proposes a new type of FFT-based spectrum plasma wave receiver that overcomes the disadvantages of conventional receivers. The receiver measures and calculates the whole spectrum by dividing the observation frequency range into three bands: bands 1, 2, and 3, which span 1 Hz to 1 kHz, 1 to 10 kHz, and 10 to 100 kHz, respectively. To reduce the size of the receiver, its analog section was realized using application-specific integrated circuit (ASIC) technology, and an ASIC chip was successfully developed. The dimensions of the analog circuits were 4.21 mm x 1.16 mm. To confirm the performance of the ASIC, a test system for the receiver was developed using the ASIC, an analog-to-digital converter, and a personal computer. The frequency resolutions for bands 1, 2, and 3 were 3.2, 32, and 320 Hz respectively, and the average time resolution was 384 ms. These frequency and time resolutions are superior to those of conventional FFT-based receivers.

Analog Circuit Failure Analysis Using Time-Resolved Emission

2005 ◽  
Author(s):  
B.J. Cain ◽  
G.L. Woods ◽  
A. Syed ◽  
R. Herlein ◽  
Toshihiro Nomura
Keyword(s):  
Analog Circuits ◽  
Integrated Circuit ◽  
Analog Circuit ◽  
Photon Emission ◽  
Nonlinear Process ◽  
Time Resolved ◽  
Non Invasive ◽  
Highly Nonlinear ◽  
Excellent Spatial Resolution ◽  
5 Ghz

Abstract Time-Resolved Emission (TRE) is a popular technique for non-invasive acquisition of time-domain waveforms from active nodes through the backside of an integrated circuit. [1] State-of-the art TRE systems offer high bandwidths (> 5 GHz), excellent spatial resolution (0.25um), and complete visibility of all nodes on the chip. TRE waveforms are typically used for detecting incorrect signal levels, race conditions, and/or timing faults with resolution of a few ps. However, extracting the exact voltage behavior from a TRE waveform is usually difficult because dynamic photon emission is a highly nonlinear process. This has limited the perceived utility of TRE in diagnosing analog circuits. In this paper, we demonstrate extraction of voltage waveforms in passing and failing conditions from a small-swing, differential logic circuit. The voltage waveforms obtained were crucial in corroborating a theory for some failures inside an 0.18um ASIC.

Layout Symmetry Annotation for Analog Circuits with Graph Neural Networks

2021 ◽  
Author(s):  
Xiaohan Gao ◽  
Chenhui Deng ◽  
Mingjie Liu ◽  
Zhiru Zhang ◽  
David Z. Pan ◽  
...  
Keyword(s):  
Neural Networks ◽  
Analog Circuits ◽  
Graph Neural Networks

scholarly journals Simultaneous Voltage and Current Measurement Instrumentation Amplifier for ECG and PPG Monitoring

Electronics ◽  
2021 ◽  
Vol 10 (6) ◽  
pp. 679
Author(s):  
Jongpal Kim
Keyword(s):  
Integrated Circuit ◽  
Cardiovascular Health ◽  
Current Noise ◽  
Instrumentation Amplifier ◽  
Current Sensing ◽  
Application Specific Integrated Circuit ◽  
Time Division ◽  
Electrocardiogram Ecg ◽  
First Time ◽  
Application Specific

An instrumentation amplifier (IA) capable of sensing both voltage and current at the same time has been introduced and applied to electrocardiogram (ECG) and photoplethysmogram (PPG) measurements for cardiovascular health monitoring applications. The proposed IA can switch between the voltage and current sensing configurations in a time–division manner faster than the ECG and PPG bandwidths. The application-specific integrated circuit (ASIC) of the proposed circuit design was implemented using 180 nm CMOS fabrication technology. Input-referred voltage noise and current noise were measured as 3.9 µVrms and 172 pArms, respectively, and power consumption was measured as 34.9 µA. In the current sensing configuration, a current noise reduction technique is applied, which was confirmed to be a 25 times improvement over the previous version. Using a single IA, ECG and PPG can be monitored in the form of separated ECG and PPG signals. In addition, for the first time, a merged ECG/PPG signal is acquired, which has features of both ECG and PPG peaks.

Neural networks integrated circuit with switchable gait pattern for insect-type microrobot

2021 ◽  
Author(s):  
Mika Kurosawa ◽  
Takuro Sasaki ◽  
Yu Usami ◽  
Shinya Kato ◽  
Arisa Sakaki ◽  
...  
Keyword(s):  
Neural Networks ◽  
Integrated Circuit ◽  
Gait Pattern

CONCEPTS OF ELECTRONICALLY-CONTROLLED ELECTROMECHANICAL/MECHANOELECTRICAL STEER-, AUTODRIVE- AND AUTOABSORBABLE WHEELS FOR ENVIRONMENTALLY-FRIENDLY TRI-MODE SUPERCARS

1994 ◽  
Vol 04 (04) ◽  
pp. 501-516 ◽  
Author(s):  
BOGDAN T. FIJALKOWSKI ◽  
JAN W. KROSNICKI
Keyword(s):  
Direct Current ◽  
Integrated Circuit ◽  
Environmentally Friendly ◽  
Alternating Current ◽  
Switching Frequency ◽  
Neuro Fuzzy ◽  
Interior Permanent Magnet ◽  
Application Specific Integrated Circuit ◽  
Computer Processor ◽  
Application Specific

Concepts of the electronically-controlled electromechanical/mechanoelectrical Steer-, Autodrive- and Autoabsorbable Wheels (SA2W) with their brushless Alternating Current-to-Alternating Current (AC-AC), Alternating Current-to-Direct Current-Alternating Current (AC-DC-AC) and/or Direct Current-to-Alternating Current (DC-AC)/Alternating Current-to-Direct Current (AC-DC) macroelectronic converter commutator (macro-commutator) wheel-hub motors/generators with the Application Specific Integrated Matrixer (ASIM) macroelectronic converter commutators (ASIM macrocommutators) and Application Specific Integrated Circuit (ASIC) microelectronic Neuro-Fuzzy (NF) computer (processor) controllers (ASIC NF microcontrollers) for environmentally-friendly tri-mode supercars (advanced ultralight hybrids) have been conceived by the first author and designed by both authors with the Cracow University of Technology’s Automotive Mechatronics Research and Development (R&D) Team. These electromechanical/mechanoelectrical wheel-hub motors/generators, respectively, for instance, can be composed of the outer rotor with the Interior Permanent Magnet (IPM) poles and the inner stator that has the three-phase armature winding. The macroelectronic converter commutator establishes the AC-AC cycloconverter, AC-DC rectifier-DC-AC inverter and/or DC-AC inverter/AC-DC rectifier ASIM macrocommutator. The microelectronic NF computer (processor) controller establishes the ASIC microcomputer-based NF microcontroller. By adopting continuous semiconductor bipolar electrical valves in the high-power ASIM, it has been able to increase the commutation (switching) frequency and reduce harmonic losses of the electromechanical/mechanoelectrical wheel-hub motors/generators, respectively.

scholarly journals Towards Amazon Forest Restoration: Automatic Detection of Species from UAV Imagery

2021 ◽  
Vol 13 (13) ◽  
pp. 2627
Author(s):  
Marks Melo Moura ◽  
Luiz Eduardo Soares de Oliveira ◽  
Carlos Roberto Sanquetta ◽  
Alexis Bastos ◽  
Midhun Mohan ◽  
...  
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Forest Restoration ◽  
Amazon Forest ◽  
Forest Species ◽  
Accuracy And Precision ◽  
Identification Of Species ◽  
Aerial Vehicle ◽  
High Resolution Images ◽  
Uav Images

Precise assessments of forest species’ composition help analyze biodiversity patterns, estimate wood stocks, and improve carbon stock estimates. Therefore, the objective of this work was to evaluate the use of high-resolution images obtained from Unmanned Aerial Vehicle (UAV) for the identification of forest species in areas of forest regeneration in the Amazon. For this purpose, convolutional neural networks (CNN) were trained using the Keras–Tensorflow package with the faster_rcnn_inception_v2_pets model. Samples of six forest species were used to train CNN. From these, attempts were made with the number of thresholds, which is the cutoff value of the function; any value below this output is considered 0, and values above are treated as an output 1; that is, values above the value stipulated in the Threshold are considered as identified species. The results showed that the reduction in the threshold decreases the accuracy of identification, as well as the overlap of the polygons of species identification. However, in comparison with the data collected in the field, it was observed that there exists a high correlation between the trees identified by the CNN and those observed in the plots. The statistical metrics used to validate the classification results showed that CNN are able to identify species with accuracy above 90%. Based on our results, which demonstrate good accuracy and precision in the identification of species, we conclude that convolutional neural networks are an effective tool in classifying objects from UAV images.

Sign in / Sign up

Export Citation Format

Share Document