Simulator for self-forming networks of informal neurons

FIELD: medicine.

SUBSTANCE: simulator comprises a number of neuron-like elements each of which contains synaptic weight changing units, a summation unit, a comparator, a converter, a random number generator, a multiplier unit, a pulse strobing unit and a prolongation unit controlling the comparator, as well as a summation unit with a regenerative loop.

EFFECT: improving the accuracy of the neural network simulation taking into account the real properties of the human and animal cerebral neurons, better understanding of the processes taking place in the prototype neuron, the use of the reproduced signal processing processes in the nerve cells to re-create artificial intelligence systems.

4 dwg

 

The technical field to which the invention relates.

The invention relates to medical Cybernetics, bionics, Neuroinformatics, in particular to the development of network systems for modeling physiological processes, as well as the creation of information-control devices and biotechnological complexes motor regulation.

The level of technology

Analogue of the invention is a device for simulation of neural networks, which is a group of synaptic interconnected formal neurons that mimic the properties of real (biological) nervous system [1, 2]. Each of the formal neurons in the network has a set of unidirectional input connections (synapses), and also has an output connection (axon), through which the signal from it goes to the synapses of the following neurons or to the input of actuator elements (analogues of motor units). Each synaptic input is multiplied by a weight corresponding to the value of the efficiency of the synapse, then all works are summed and fed to one of the inputs of the comparator, the threshold is determined by the amplitude of the control signal supplied to the other input. General view of the formal neuron that implements a set of specified properties, shown in figure 1. There are many input signals denoted by the vector Hlady weight W i(1) corresponds to the efficiency of a synaptic connection. The output of the adder 2 is connected to one input of the comparator 3, the threshold characteristics are set by the signal X0supplied to another (the Trustee) of its input. Converter terminal 4 determines the generation parameters of the neuron action potentials [f(u)].

With essential features of the invention match the following set of features analog: "each of the neurons in the network has a set of unidirectional input connections (synapses), and also has an output connection (axon), through which the signal from it goes to the synapses of the following neurons or to the input of actuator elements (analogues of motor units). Each synaptic input is multiplied by a weight corresponding to the value of the efficiency of the synapse, then all works are summed and fed to one of the inputs of the comparator, the threshold is determined by the amplitude of the reference signal supplied to the other (control) his entry".

The disadvantage of analog is incomplete modeling of biological prototype, namely, the absence of a formal model of a neuron of a number of basic functional properties of real nerve cells. Special attention in this respect deserve: a stochastic (probabilistic) nature of the dependence of the pulse the biological activity of a neuron from the parameters of the incoming synaptic influx [3], the presence of functional feedbacks in the Central and peripheral synapses chemical type [4]; the existence of vnutrennih (endogenous) mechanisms of generation of action potentials [5]; operant determinacy of the process of formation of endogenous patterns of neural activity [6]; the role of the total activity of the system in the determination of the functions of its individual elements [7].

The prototype of the invention is a Device for neuron simulation" (Pruhonice Y.A. Author. mon. N1709356 A1 (SU) G06G 7/60 (1989)) [8], which can improve the accuracy of the modeling of neural networks by representing processes going on in them as variables stochastic sequences.

With essential features of the invention match the following set of features of the prototype: "in the device containing the blocks changes of synaptic weights, the first inputs of which are the inputs of the device, the spatial adder, the inputs of which are connected to the outputs of the blocks changes of synaptic weights entered ... the comparator logic block and the random number generator".

The disadvantage of the prototype is also a mismatch between the number of essential characteristics of the fundamental principles of the functioning of real neurons and their networks. First of all, it ignorirovaniya role of the total activity multineuron systems in the processes of their formation [7] and the exception mechanism of synaptic feedback [4].

Flaw remediation can be achieved due to the fact that at the input of the threshold device (comparator) signal is not output from the adder of synaptic weights, and the output of the multiplier performing the function of modulating the synaptic signal stochastic stream of random numbers. Accordingly, on the other (control) input of the comparator is fed strobirovaniya the action potentials of the neuron output signal of the second (system) adder for N input of which receives the action potentials of all N neurons in the network. As a result, the output of each neuron is formed by a sequence of action potentials, the probability of generating each of which depends not only on the current settings of synaptic inflow, but also on the amplitude of the terminal activity of the entire neural network as a whole.

The invention

The problem to which the invention is directed, is to develop a biologically plausible neural network model ("informal network of neurons"), taking into account the whole complex is significantly important functional characteristics of biological prototype.

During implementation of the invention can be obtained the following results:

A. improving the accuracy of neural network modeling taking into account the real properties of neurons in the brain of humans and animals.

B. the Best is animania processes, what is happening in the neuron prototype.

C. Use reproduced by processing signals in the neuron for the creation of artificial intelligent systems.

These technical results are achieved due to the fact that the composition of the known device containing blocks changes of synaptic weights 1, an adder 2, a comparator 3, the inverter 4, the random number generator 5, enter the multiplier 6 and the second (system) the adder 9, the N inputs of which the output signals of the N neurons (Posa), and the output of the adder via individual for each neuron unit pulse Gating 7 and the block extension 8 connected to the control input of the comparator 3 (figure 2).

The features that distinguish the invention from the closest analogue, expressed by the following set of features: "the known device enter the multiplier and the second (system) adder, N inputs of which the output signals N informal neurons, and the output of the adder through the block pulse Gating and block extension connected to the control input of the comparator".

With the aim of improving the efficiency of modeling, as well as the most accurate reflection in the proposed model the main features of its biological prototype, the possibility of selective and generalized Shin is pricescope impact on all neurons in the network.

Thus, the objective of the invention is resolved.

Information confirming the possibility of carrying out the invention

The possibility of carrying out the invention is confirmed by the results of multiparameter computer analysis [9], conducted in the environment of object-oriented modeling LabView 2010.

The complete block diagram of the device for simulation of self-organizing informal networks of neurons in the LabView environment is presented in figure 3. The scheme consists of 4 main sectors. The first of them (Fig) brings together twenty-informal neurons 1, the output signal of each of which is fed to the input of the Central accumulator 2 through the inverter 3. The second sector block diagram (Fig) is a device forming external ("synaptic") control signal probability of generating a neuron action potentials. The main element here is "Simulate Sig" 4. His work through a positive feedback loop, going from the Central exit of the adder 2 manages virtual subprior "Select" 5. Submission on its Central (control) input logical unit causes it to disconnect from the zero point (OffSet 3) and switching with digital DC level of synaptic inflow (t).

The function of the third sector (Fig) is the automatic shutdown of all systems is after it has reached its result (i.e. the after forming the output of the Central adder 2 a certain pattern of efferent impulses). The main functional element here is the RS-trigger, collected on two Boolean elements "And Not" 6. Activate "readiness" is a Boolean signal coming from the switch 7. The transition of the RS-flip-flop in an alternative stable state is produced by the pulse received from the comparator 8. And finally, the fourth sector (Fig) is the amplifier-shaper sound signals corresponding to the output pulses of the whole system. Implementation of this function is two built-in LabView by subpriority: Simulate Sig 9 and Play Waveform 10.

The device operates as follows.

In each iteration, each informal neuron network 1 (figure 3. 1), total internal scheme is presented in figure 4, is formed one random number and, accordingly, is one action compare this number with the value of the control signal. In that case, if the random number is less than or equal to, the output of the comparator appears a short pulse of logical units. Otherwise, its output remains a logical zero. For the formation of a stochastic sequence of pulses is used, the stream generator of random numbers double precision {Random Number} 2, the value 0 is about 1 (figure 4). The output of the Random generator 2 through the tube 3 is connected to the comparator 4 {Less or Equal}to another entrance through which the block Select 5 filed managing (threshold) signal.

After starting the device, in some moments in the result of a coincidence in time of generation of action potentials from different neurons in the network at its output appear total amplitude pulses. Automatic registration of the moments of their appearance and positive "reinforcement" creating these numeric anomalies of the elements of the system by using logical block 6. The essence of his work is that each neuron network each generated pulse not only brings its own micro contribution to the total output of the system, but every time gates ("contractor") the current state of this output. As a result, the occurrence of any high-amplitude fluctuations is logged only those elements, the pulse activity which has led to its formation.

When determining the mode of feedback was taken into account the results of the experiments on the operant conditioning bit of activity of individual nerve cells [6]. In this device (figure 4) implementation of the principle of operant determination was carried out as follows. To reach the amplitude of the output signal upper threshold value is of 8 the system operates in its normal mode of generation of the stochastic stream of pulses. In the case of any excess collected on two elements "And Not" RS-trigger 6 is transferred by the action potential of a neuron in an alternative stable state. As a result, the control input of the respective comparator, instead of continuously varying signal total output of the system is extended ("hangs") of its peak level, which is maintained until after the generation of the next neuron action potential, but at a low level 7 network activity. The simultaneous development of these processes in different channels of the device leads to forming system elements move in the cooperative mode of coordinated activity. On the level of the input link of a neural network is manifest in the emergence of a functional feedback, in which the pulse output of informal neurons begins to actively influence the weight options available to them external signals.

The possibility of carrying out the invention with the implementation of this assignment is confirmed by the known means and methods of receiving devices. The results of testing the models considered above, as well as other developed on the basis of the instruments presented in the monograph "fundamentals of quantum synergetics functional systems, part II" (ISBN 078-5-91506-026-4. Moscow. 2011 author: beaver L.V.).

So the m way confirmed the possibility of carrying out the invention.

Technical results of the invention are: to improve the accuracy of neural network modeling taking into account the real properties of neurons in the brain of humans and animals; a better understanding of the processes in the neuron prototype, and the use of replicated processing of signals in nerve cells to create artificial intelligent systems.

Sources of information

1. Mac-Calloc U., Pitts Century Logical calculus of the ideas related to neural activity // Neural networks: history of development theory. Edited Galushkina A.I., Tsypkin AS IPIR. 2001. Pp.5-22.

2. Red'ko V.G., yanushevski I.A. Formal neuron. Auth. mon. N1013984 A (SU) 00667/60 (1983).

3. Burns B.D. The uncertain nervous system. 1968. London. Ed.: E.Arnold. 263 p.

4. Matyushkin D.P. Feedback synapse. Leningrad: Nauka. 1989.

5. Pacemaker potential of the neuron. Ed. by E.N. Sokolov. Tbilisi. 1975.

6. Stein L, Belluzzi JD Operant conditioning of individual neurons. In: M.L.Commons, R.M.Church, J.R.Stellar, A.R.Wagner (Eds.). Quantitative analyses of behavior. V.7. Biological determinants of reinforcement and memory 1988. (pp.249-264). Hillsdale, NJ: Erlbaum.

7. Anokhin P.K. System analysis of the integrative activity of the neuron - Advances in physiological Sciences. 1974. V.5. No. 2. C.5-92.

8. Pruhonice Y.A. Device for simulation of a neuron. Auth. St. N1709356 A1 (SU) G06G 7/60 (1989).

9. Bobrovnikov L.V. foundations of quantum synergetics function of the national systems, part II". ISBN 078-5-91506-026-4. Moscow. 2011. 162 C.

Device for modeling self-organizing informal networks of neurons, which includes the set of neuron-like elements, each of which contains blocks changes of synaptic weights, the adder, a comparator, an inverter and a random number generator, characterized in that the composition of each network element includes a multiplier, which modulates synaptic signal pulses of the generator of random numbers and sends formed in this way, the stochastic pulse stream input common for all network neurons adder with a positive feedback loop through which the total output of the system after its individual Gating the action potentials of individual neurons with subsequent programmable extension distributed in managing their inputs of Comparators.



 

Same patents:

FIELD: information technology.

SUBSTANCE: device includes a plurality of neuron-like base elements (BE) which are grouped into modular units such that outputs of like BE are orthogonal to inputs of other BE and are separated by a dielectric film with a semiconductor layer. The neuron-like BE has a multipoint input for reception and summation of electrical signals, a working output and a signal processing unit, having a threshold voltage former, a comparator and two normalised voltage formers.

EFFECT: simple design of the device, broader functional capabilities thereof and a more complex category of problems solved.

10 cl, 7 dwg

Neuroprocessor // 2473126

FIELD: information technology.

SUBSTANCE: invention can be used in designing computing means for systems for controlling highly manoeuvrable aviation and space-rocket objects where there is need for fast computation of functions, for example, trigonometric functions, used in matrix transformations when solving tasks of forming an inertial coordinate system based on information from angular velocity sensors, and when solving tasks for maintaining operating capacity of computers during changes in parameters of LSI elements due to the effect of natural or artificial ionising radiations. The device has a unit for communication with an on-board computer or which controls a higher level subsystem, a control device, a buffer register, a memory device, multipliers, an adder and an output register.

EFFECT: fewer faults in microchips.

5 cl, 5 dwg

FIELD: physics.

SUBSTANCE: method comprises steps for: obtaining seismic data for an area of interest; obtaining an initial seismic cube using said seismic data, wherein the initial seismic cube is a three-dimensional representation of the seismic data; generating a plurality of shifted seismic cubes within the area of interest using said seismic data and a shifting parameter, wherein each of the plurality of shifted seismic cubes is shifted from the initial seismic cube; and wherein the shifting parameter defines a direction and a range that the initial seismic cube should be shifted; generating a neural network using the initial seismic cube, the plurality of shifted seismic cubes, and well log data; and applying the neural network to said seismic data to obtain a model for the area of interest, the model being configured for use in adjusting an operation of the wellsite.

EFFECT: high accuracy.

20 cl, 19 dwg

FIELD: information technology.

SUBSTANCE: in an ophthalmic-microsurgical computer local area network for vitreoretinal operations, formatting devices are in form of a radial-annular structure consisting of a single set of automated workstations (AWS), which synchronously or asynchronously functioning, processing, converting, transmitting, analysing, synthesising hierarchical structures of an artificial neural network: diagnosis AWS (DAWS), ophthalmic-microsurgical AWS (OMAWS), subsequent operation stages AWS (SOSAWS), component AWS (CAWS), surgeon's operating unit (SOUAWS), with opposite forward and reverse flow of information in between, where each AWS has at least one neural circuit, interconnected identification units (IU), an interpolation unit (INU), an extrapolation unit (EU), which are the neural network converting and transmitting elements (NNCTE), a decision unit (DU), which is the neural network analysis and synthesis element (NNASE).

EFFECT: simultaneous improvement of accuracy of determination and quality of identifying diagnoses, determining indications for conducting operations, high selectivity when conducting operations, accuracy in determining the sequence of operations, simulating operations, accuracy in choosing the anaesthetic method, accuracy of providing implants and expendable materials, optimisation of flow of information and necessities during vitreoretinal ophthalmic-microsurgical operations.

1 dwg

FIELD: physics.

SUBSTANCE: neuron simulation method is based on calculation of squares of Euclidean distance from the input vector to each of 2n vertices of a unit n-dimensional cube in weighting units, and multiplication of values inverse to these distance values with components of the target vector respectively, and then summation in an adder and conversion in the activation unit through an activation function.

EFFECT: possibility of simulating a neuron of any given Boolean function from a complete set of from n variables.

6 dwg, 1 tbl

FIELD: information technology.

SUBSTANCE: multilayer modular computer system has several layers, including a neural network layer, a transport layer and a processor layer, wherein the transport layer contains network controller-router modules, the processor layer contains processor modules, and all the said modules have multiple inputs and outputs connected to each other and connected to the inputs and outputs of the system. The processor modules train neural network domain modules.

EFFECT: high decision speed, possibility of grafting layers and modules in each layer during operation of the system with a complex task, high reliability of the computer system.

3 cl, 1 dwg

FIELD: information technologies.

SUBSTANCE: invention may be used for building of modular neural computers, which function in symmetrical system of residual classes. Stated neuron network comprises unit of neuron network of end ring of senior coefficient generation for generalised positional system of numeration, unit of polarity shift, unit of error detection, buses "with errors" and "without errors".

EFFECT: reduced hardware complexity.

3 dwg

FIELD: physics; computer engineering.

SUBSTANCE: present invention pertains to neurocomputers. The device has a unit for storing a binary input signal, a logic AND-OR circuit, internal memory unit, unit for generating the output string of codes, a generator of synchronising pulses, control unit, a unit for selecting duration and extracting information, analysis block and a corrector unit.

EFFECT: increased rate of operation, providing for the possibility of distinguishing change in state of processed signals, increased noise immunity, possibility of making super-complex neural networks, and simplification of design.

9 cl, 1 dwg

FIELD: modular neuro-computing systems.

SUBSTANCE: neuron network contains input layer of neurons, at inputs of which residuals of number being divided are received through system of modules, (n-1) neuron networks of finite ring for addition, (n-1) neuron networks of finite ring for multiplication, neuron network for expanding a tuple of numerical system of residues, and as output of neuron network for dividing numbers represented in system of residual classes are outputs of neuron network of finite ring for multiplication and output of neuron network for expansion of tuple of numerical system of residues.

EFFECT: expanded functional capabilities, increased speed of division, reduced volume of equipment.

1 dwg

FIELD: neuron-like computing structures, possible use as processor for high speed computer systems.

SUBSTANCE: device contains artificial neuron network composed of analog neurons, at least one controllable voltage block, a group of long neuron-like nonlinear communication units, each one of which contains serially connected circuit for synchronization and selection of radio impulse envelope, auto-generator with self-suppression circuit, a length of coaxial line, realizing functions of antenna, additional circuit for synchronization and selection of radio-impulse envelope.

EFFECT: increased information processing speed due to increased paralleling degree of computing processes.

2 dwg

Neuron-like element // 2295769

FIELD: cybernetics, possible use as a cell for neuron networks.

SUBSTANCE: neuron-like element may be used for realization on its basis of neuron network for solving problems of estimation of functioning of complicated open systems, estimation of degree of optimality of various solutions by ensuring possible construction of model of researched system, both hierarchical and recurrent, with consideration of varying original and working condition of its elements and variants of their functioning, during modeling taking into consideration the level of self-sufficiency of neuron-like elements, susceptibility to effect of external signals, type and errors of setting of their parameters and parameters of input signals, and also provision of given precision of self-teaching of neuron network. Device contains input block, block for setting and normalizing weight coefficients, block for computing parameters of input signals, adder, signals share limiter, block for computing input part of condition, block for setting internal state, block for computing internal part of distance, block for counting distance, memory block, analyzer of state change value, block for determining precision of self-teaching of neuron network, block of determined dependencies, switch, output block, control block, random numbers generator.

EFFECT: creation of neuron-like element.

2 cl, 1 dwg

FIELD: computer engineering, possible use in modular neuro-computer systems.

SUBSTANCE: in accordance to invention, neuron network contains input layer, neuron nets of finite ring for determining errors syndrome, memory block for storing constants, neuron nets for computing correct result and OR element for determining whether an error is present.

EFFECT: increased error correction speed, decreased amount of equipment, expanded functional capabilities.

1 dwg, 3 tbl

FIELD: neuro-cybernetics, possible use in artificial neuron networks for solving various problems of logical processing of binary data.

SUBSTANCE: method for realization of logical nonequivalence function by neuron with two inputs is based on multiplication of input signals with corresponding weight coefficients and summing them, after that the total is transformed in activation block firstly by quadratic transfer function, and then by threshold function at neuron output.

EFFECT: realization by one neuron of first order of logical nonequivalence function of two variables.

5 dwg, 1 tbl

FIELD: neuron-like computing structures, possible use as processor for high speed computer systems.

SUBSTANCE: device contains artificial neuron network composed of analog neurons, at least one controllable voltage block, a group of long neuron-like nonlinear communication units, each one of which contains serially connected circuit for synchronization and selection of radio impulse envelope, auto-generator with self-suppression circuit, a length of coaxial line, realizing functions of antenna, additional circuit for synchronization and selection of radio-impulse envelope.

EFFECT: increased information processing speed due to increased paralleling degree of computing processes.

2 dwg

FIELD: modular neuro-computing systems.

SUBSTANCE: neuron network contains input layer of neurons, at inputs of which residuals of number being divided are received through system of modules, (n-1) neuron networks of finite ring for addition, (n-1) neuron networks of finite ring for multiplication, neuron network for expanding a tuple of numerical system of residues, and as output of neuron network for dividing numbers represented in system of residual classes are outputs of neuron network of finite ring for multiplication and output of neuron network for expansion of tuple of numerical system of residues.

EFFECT: expanded functional capabilities, increased speed of division, reduced volume of equipment.

1 dwg

FIELD: physics; computer engineering.

SUBSTANCE: present invention pertains to neurocomputers. The device has a unit for storing a binary input signal, a logic AND-OR circuit, internal memory unit, unit for generating the output string of codes, a generator of synchronising pulses, control unit, a unit for selecting duration and extracting information, analysis block and a corrector unit.

EFFECT: increased rate of operation, providing for the possibility of distinguishing change in state of processed signals, increased noise immunity, possibility of making super-complex neural networks, and simplification of design.

9 cl, 1 dwg

FIELD: information technologies.

SUBSTANCE: invention may be used for building of modular neural computers, which function in symmetrical system of residual classes. Stated neuron network comprises unit of neuron network of end ring of senior coefficient generation for generalised positional system of numeration, unit of polarity shift, unit of error detection, buses "with errors" and "without errors".

EFFECT: reduced hardware complexity.

3 dwg

FIELD: information technology.

SUBSTANCE: multilayer modular computer system has several layers, including a neural network layer, a transport layer and a processor layer, wherein the transport layer contains network controller-router modules, the processor layer contains processor modules, and all the said modules have multiple inputs and outputs connected to each other and connected to the inputs and outputs of the system. The processor modules train neural network domain modules.

EFFECT: high decision speed, possibility of grafting layers and modules in each layer during operation of the system with a complex task, high reliability of the computer system.

3 cl, 1 dwg

FIELD: physics.

SUBSTANCE: neuron simulation method is based on calculation of squares of Euclidean distance from the input vector to each of 2n vertices of a unit n-dimensional cube in weighting units, and multiplication of values inverse to these distance values with components of the target vector respectively, and then summation in an adder and conversion in the activation unit through an activation function.

EFFECT: possibility of simulating a neuron of any given Boolean function from a complete set of from n variables.

6 dwg, 1 tbl

FIELD: information technology.

SUBSTANCE: in an ophthalmic-microsurgical computer local area network for vitreoretinal operations, formatting devices are in form of a radial-annular structure consisting of a single set of automated workstations (AWS), which synchronously or asynchronously functioning, processing, converting, transmitting, analysing, synthesising hierarchical structures of an artificial neural network: diagnosis AWS (DAWS), ophthalmic-microsurgical AWS (OMAWS), subsequent operation stages AWS (SOSAWS), component AWS (CAWS), surgeon's operating unit (SOUAWS), with opposite forward and reverse flow of information in between, where each AWS has at least one neural circuit, interconnected identification units (IU), an interpolation unit (INU), an extrapolation unit (EU), which are the neural network converting and transmitting elements (NNCTE), a decision unit (DU), which is the neural network analysis and synthesis element (NNASE).

EFFECT: simultaneous improvement of accuracy of determination and quality of identifying diagnoses, determining indications for conducting operations, high selectivity when conducting operations, accuracy in determining the sequence of operations, simulating operations, accuracy in choosing the anaesthetic method, accuracy of providing implants and expendable materials, optimisation of flow of information and necessities during vitreoretinal ophthalmic-microsurgical operations.

1 dwg

Up!