The paper describes an approach for implementing Boolean neuralnetworks on silicon. The hardware is based on a custom designed Field Programmable Logic Device (FPLD) which integrates 'synapses' and 'neuro...
详细信息
The paper describes an approach for implementing Boolean neuralnetworks on silicon. The hardware is based on a custom designed Field Programmable Logic Device (FPLD) which integrates 'synapses' and 'neurons' and allows random access to the weights during training. networks are realised from arrays of the neural chip which are assembled on ceramic as Multi-Chip Modules (MCM) to provide expandability and flexibility. The hardware provides parallel computation of the 'neuron' outputs and promises significantly improved performance compared to purely software approaches.
The introduction of transmission system based on the Synchronous Digital Hierarchy (SDH) has created new opportunities for flexible and resilient transmission networks. This paper looks at reasons to deploy SDH in rai...
详细信息
The introduction of transmission system based on the Synchronous Digital Hierarchy (SDH) has created new opportunities for flexible and resilient transmission networks. This paper looks at reasons to deploy SDH in railway networks, and reviews some of the design considerations. The complementary technologies of WDM, ATM and flexible adaptation multiplexers are also introduced. Some of the new services which will become available once broadband capacity can be provisioned from the network to stations, control centres and the railway lineside are described.
The Advanced Computer Architecture Group at York has initiated the Advanced Uncertain Reasoning Architecture (AURA) project which aims to develop a new class of architecture for integrated symbolic and sub-symbolic pr...
详细信息
The Advanced Computer Architecture Group at York has initiated the Advanced Uncertain Reasoning Architecture (AURA) project which aims to develop a new class of architecture for integrated symbolic and sub-symbolic processing. These architectures will be based on a class of neural network called binary correlation matrix memory (CMM). The approach adopted supports optimal associative memory mappings, which are robust in the presence of noisy or imperfect inputs, and offer considerable flexibility in the representation of knowledge. An example of how AURA can provide anytime functionality in real time identification has been described.
The reconfigurability of certain field programmable gate arrays(FPGAs) has shown their advantage of flexibility in digital system design. With the availability of greater density and high speed of FPGAs, the ability t...
详细信息
The reconfigurability of certain field programmable gate arrays(FPGAs) has shown their advantage of flexibility in digital system design. With the availability of greater density and high speed of FPGAs, the ability to realise special purpose processors will become possible. In this paper, we present our research work of implementing specific trained neural network in Xilinx XC4000 series FPGAs for portable digital system prototype in heart disease classification process.
There have been several studies of neuralnetworks whose response properties approximate those found in mammalian visual systems. Most of these have been concerned with the evolution of single units or small sets of u...
详细信息
There have been several studies of neuralnetworks whose response properties approximate those found in mammalian visual systems. Most of these have been concerned with the evolution of single units or small sets of units having receptive field profiles which are eigenvectors of an appropriately defined energy function. While studies of individual units are important, one of the most striking features of the visual system is its global structure, represented by the so called retinotopic maps, found at all levels of the system studied so far. The probable structure and perceptual function of such a visual system are discussed.
As is well known, Flexible AC Transmission systems (FACTS) provide opportunities of better utilizing existing transmission systems by using power electronics based controllers. One of the main FACTS devices is control...
详细信息
ISBN:
(纸本)0852965699
As is well known, Flexible AC Transmission systems (FACTS) provide opportunities of better utilizing existing transmission systems by using power electronics based controllers. One of the main FACTS devices is controllable Series Compensation (CSC), which has an ability to control the compensated impedance by changing the firing angle of thyristor. However, the implementation of this technology will pose new problems to the conventional line protection schemes. This paper proposes a novel adaptive protection schemes for CSC transmission systems by using a neural network approach. Based on the author's previously comprehensive digital simulation studies of CSC transmission systems, this paper places emphasis on the feature extraction the topology and training of neutral networks. Some preliminary test results clearly show the trained neural network is able to make correct trip decision from abnormal voltage waveforms using associations learned from previous experiences. In addition, the scheme also has the ability to identify faulted phase. The test results successfully demonstrate the feasibility of the neuralnetworks based adaptive protection for the CAS transmission systems.
The proceedings contains 7 papers on real time knowledge based systems. Topics discussed include fault diagnosis using cost bounded possibilistic Assumption based Truth Maintenance System (ATMS), scheduling for time c...
详细信息
The proceedings contains 7 papers on real time knowledge based systems. Topics discussed include fault diagnosis using cost bounded possibilistic Assumption based Truth Maintenance System (ATMS), scheduling for time constrained model based diagnosis, neuralnetworks for real time knowledge based applications, methodology for developing model based diagnostic systems, large scale real time knowledge based methods and technology, bloom and billet mill expert systems, and a model based and rule based real time tool for monitoring and diagnosis of dynamic systems.
The competence of the HyperNet system was evaluated. In addition, an advanced path able to satisfy the standards of more complicated, extensive, real-time uses were discussed. The HyperNet system integrates probabilis...
详细信息
The competence of the HyperNet system was evaluated. In addition, an advanced path able to satisfy the standards of more complicated, extensive, real-time uses were discussed. The HyperNet system integrates probabilistic, RAM-based, feed forward design and employs a conventional VLSI IC with on-board Reward-Penalty learning program. The system also provides opportunity for inexpensive Ultra Large Scale and Wafer Scale Integration.
A VLSI Implementation of an Artificial neural Network using a single n-channel MOS transistor per synapse is investigated. The simplicity of the design is achieved by using pulse width modulation to represent neural a...
详细信息
A VLSI Implementation of an Artificial neural Network using a single n-channel MOS transistor per synapse is investigated. The simplicity of the design is achieved by using pulse width modulation to represent neural activity and a novel technique for manipulating synaptic weights. A multi layer perceptron network built in hardware gives good results for a simple classification task.
The general principles of neural and hybrid architectures for multimedia in general are discussed. From the perspective of knowledge engineering, hybrid symbolic/neural agents are advantageous since different mutually...
详细信息
The general principles of neural and hybrid architectures for multimedia in general are discussed. From the perspective of knowledge engineering, hybrid symbolic/neural agents are advantageous since different mutually complementary properties can be combined. Symbolic representations have advantages with respect to easy interpretation, explicit control, fast initial coding, dynamic variable binding and knowledge abstraction. neural agents show advantages for gradual analog plausibility, learning, robust fault-tolerant processing, and generalization to similar input. Since these advantages are mutually complementary, a hybrid symbolic neural architecture can be useful if different processing strategies have to be supported.
暂无评论