Abstract: In a control method and a control device for an autonomously exploring robot, sensory values induce, in a strategy generating network, control sets according to inner connectivity which is acquired in a training phase. The control sets control acting units which act on the environment. Control sets and sensory values are given as input to a network system which distinguishes between already trained pairs of values and unknown pairs of values.

Abstract: A plurality of programs are generated which are each composed of an arrangement of basic instructions of plural types required to generate output data with respect to given input data of given types. Each of the programs is executed with a computer until a predetermined condition is satisfied, and the input data is given to the computer. Output data generated with respect to the input data when each of the programs is executed is compared with reference output data predetermined with respect to the input data, and an energy value indicative of the quality level of each of the programs is incremented or decremented such that the energy value is greater as the degree with which the generated output data and the reference output data agree with each other is higher. Each of the programs is copied or eliminated depending on the magnitude of the energy value after the execution of each of the programs by the computer is finished, and the programs are crossed and mutated with a predetermined probability.

Abstract: A method and system electronically fuzzy classify a pixel belonging to a set of digital image data with respect to a membership of the pixel in a plurality of image classes. This process determines a fuzzy classification of the pixel and generates an effect tag for the pixel based on the fuzzy classification determination. Each class is defined by a set of heuristic rules such that the image classes are non-mutually exclusive. The heuristic rules are a set of conditions that define the membership value of the pixel within a certain class, thereby allowing a pixel to have a membership value in every possible image class.

Abstract: A method of prefetching data from the storage media of a data storage system, in which the data is stored on, and read from, the storage media in individually-accessible data storage portions, such as tracks of a hard disk. A number of immediately past-read data storage portions are used to predict a storage portion or portions that will likely be requested in the future. The unique identifiers of those previously-read data storage portions are determined, and a prediction coefficient is calculated for each unique identifier. The prediction coefficient is then multiplied by the unique identifier, and the results are summed to determine the unique identifier of a data portion to be prefetched.

Abstract: There is provided by the present invention a database of cases in which index-labeled case data is accumulated, an index appearance matrix generator for generating an index appearance matrix as a target according to a first index from the case data an index appearance matrix generator for generating a index appearance matrix as a reference according to a second index from the case data, a cross index matrix generator for generating a cross index matrix from these index appearance matrixes, an inter-index attribute value fuzzy relation computing section for computing a fuzzy relation between index attribute values from the cross index matrix, and a fuzzy thesaurus generating section for generating a fuzzy thesaurus from a fuzzy relation between the index attribute values.

Abstract: A system controller which significantly reduces the number of development processes and automatically exercises control corresponding to situations. The system controller is made up of a manipulation value output circuit for supplying a manipulation value to a system to be controlled, a control example memory for storing the manipulation value and a controlled variable which is output from the system in response to the manipulation value, a control rule extraction circuit for extracting a control rule from among a plurality of control examples stored in the control example memory, and a manipulation value computing circuit for calculating a manipulation value which matches the controlled variable with a target figure using the control rule extracted by the control rule extraction circuit, and causing the manipulation value output circuit to output the thus obtained manipulation value.

Abstract: An adaptive filtering neural network classifier for classifying input signals, includes a neural network and one or more adaptive filters for receiving input analog signals to be classified and generates inputs for the classifier. Each adaptive filter is characterized as having a predetermined number of operating parameters. An analog to digital converter converts each input signal into a digital signal before input to the neural network. The neural network processes each digital signal to generate therefrom a plurality of weighted output signals in accordance with the type of network implemented. One of the weighted output signals represents a class for the input signal, and an error signal representing a difference between the weighted output signals and a predetermined desired output is also generated by the network.

Abstract: A method for executing partially trusted behaviors using an automated assistant. The method employs a guarded execution mode which provides for the safe execution of partially trusted behaviors. Further, incremental trust growth is provided for allowing more behaviors to be executed autonomously by the automated assistant. Trust growth is accomplished through the use of a trust library. The method also employs trust reuse which provides for the reuse of individual trust library entries across behaviors which are proven to be erroneous by the automated assistant.

Abstract: When the system simulates changes of a virtual world and behavior of virtual creatures inhabiting it, the unit time of real time corresponding to the unit time necessary for one processing by the system is divided as time grading. Specifically, when the system is booted up, corresponding to the duration from the time the system power source is cut off previously, simulation is performed by setting the time grading in one second, one minute, one hour, one day, one week, one month, or one year, for one processing by the system. By upon starting-up of the system, introducing such concept of time grading for the time series, simulation is performed at the time grading corresponding to the duration of the period in which the system has been stopped so far, so that prompt simulation is realized.

Abstract: A robot includes a robot body to which the procedure of a task is teachable, a teaching portion, an operating portion, a control portion, and a changeover portion. The teaching portion includes a signal generating portion for generating a drive signal for driving the robot body, and a deadman switch which indicates by its opened/closed state whether the procedure of the task is being taught to the robot body or not. The operating portion generates a start signal for starting the robot body. The control portion receives the drive signal and a signal indicating the opened/closed state of the deadman from the teaching portion and the start signal from the operating portion. The changeover portion operates so that the start signal from the operating portion is prevented from being supplied to the control portion when the teaching portion is electrically connected to the control portion.

Abstract: An artificial intelligence system is provided which makes use of a dual subroutine to adapt weights. Elastic Fuzzy Logic ("ELF") System is provided in which classical neural network learning techniques are combined with fuzzy logic techniques in order to accomplish artificial intelligence tasks such as pattern recognition, expert cloning and trajectory control. The system may be implemented in a computer provided with multiplier means and storage means for storing a vector of weights to be used as multiplier factors in an apparatus for fuzzy control.

Abstract: A solder paste brick inspection and physical quality scoring system 10 employs a neural network 70 trained with a fuzzified output vector. An image of solder paste bricks 64 on a printed circuit board 12 is acquired by a CCD camera 30. Values of a predetermined set of brick metrics are extracted from the image by a computer 28 and used as a crisp input vector to trained neural network 70. A defuzzifier 76 converts a fuzzy output vector from neural network 70 into a crisp quality score output which can be used for monitoring and process control.

Abstract: A system for controlling the movement of at least one limb of a person's body includes an interface unit in communication with the person's body for receiving an input signal. A polynomial processor generates a control signal as a polynomial function of the at least one input signal. A generator generates a functional electrical stimulation signal. A plurality of electrodes stimulates the movement of the at least one limb by introducing the functional electrical stimulation signal to the at least one of a plurality of inactive muscles.

Abstract: An apparatus and method for solving problems where a population is created and evolved to generate a result. While solving the problem, the architecture of entities in the population are altered. Each of said entities may include internally and externally invoked sub-entities. The externally invoked sub-entities are capable of having actions, invocations of sub-entities which are invoked internally, and material. Also, each sub-entity which is invoked internally is capable of including actions, invocations of internally invocable sub-entities, material provided to the externally invocable sub-entity, and material.

Abstract: A behavioral based environment system and method for controlling an interactive playground. The system includes a system supervisor unit that utilizes a rule file, a scene file and a MIDI file in conjunction with a variety of sensor input to create an appropriate system response. Output control signals generated by the system supervisor unit are transmitted to other coupled computers to effectuate audio, visual and other effects in an interactive playground environment. The system supervisor has the desirable ability to load different scene, rule and MIDI files to create different system behavior in response to sensor stimuli, thereby creating a more adaptive behavioral based environment.

Abstract: A system for performing a plurality of independent analysis procedures simultaneously, each procedure having a sample and at least one process step for operating on that sample the system having a robotic device for moving a sample among a plurality of processing stations where the robotic device comprises substrate to which are coupled a plurality of tiles each one of the tiles disposed at a selected location on the substrate; a plurality of reagent trays coupled to the tiles, each one of the reagent trays having substantially identical shape and size; a robotic arm disposed for reaching substantially all the tiles; and capable of coupling and decoupling from a sample holder, the sample holder capable of holding at least one sample in a selected position, whereby the robotic arm is capable of moving the sample from a first one of the reagent trays to a second one of the reagent trays; and a processor for selecting, at a plurality of times a sample to be moved, and for directing the robotic arm to move the sa

Abstract: A rule based processing system that is optimized for implementation in hardware. The system includes a set of logic gates that are used to implement a rule network connected to a working memory composed from a set of memory components. When a set of inputs are presented to the logic gates, a set of inferences are formed and the results are stored in the memory components. The inferences that are stored in the working memory are used to determine the state of the system at any given time. These inferences can be output to an external device to report on the status of the system. The rule based production system is designed to be implementable on a high performance semiconductor micro-chip and is thus suitable for use in embedded real-time applications.

Abstract: A method and apparatus for providing access to a banked peripheral memory via a contiguous linear address space. The present invention provides a linear address space having a present region that is mapped to a host memory region of a computer system. The present invention further provides a relocatable selector that provides access to a portion of the linear address space. Accessing programs exchange data with the banked peripheral memory via the relocatable linear address space. When an accessing program references an address of the relocatable address base that is not mapped to the present region, the relocatable linear address space is positioned so that the referenced address maps to the present region. Additionally, a bank of the peripheral memory that corresponds to the referenced address is also mapped into the host memory region so as to enable the accessing program to exchange data with the banked peripheral memory via the relocatable linear address space.

Abstract: A method of extracting a gate model from a fet model using a computer implemented expert system apparatus to perform the steps of recognizing power, ground and clock signals; recognizing inverters; recognizing and preserving all logic signals of the fet modeled circuit; building one or more structurally based boolean partial trees for each of the recognized logic signals; heuristically pruning the one or more boolean trees; and building logic equations from the one or more boolean partial trees. The expert system apparatus comprises a fet modeled input netlist, an inference engine, a rule base, a user input, and a gate modeled output netlist.

Abstract: A type of neural network called a self-organizing map (SOM) is useful in pattern classification. The ability of the SOM to map the density of the input distribution is improved with two techniques. In the first technique, the SOM is improved by monitoring the frequency for which each node is the winning node, and splitting frequently winning nodes into two nodes, while eliminating infrequently winning nodes. Topological order is preserved by inserting a link between the preceding and following nodes so that such preceding and following nodes are now adjacent in the output index space. In the second technique, the SOM is trained by applying a weight correction to each node based on the frequencies of that node and its neighbors. If any of the adjacent nodes have a frequency greater than the frequency of the present node, then the weight vector of the present node is adjusted towards the highest-frequency neighboring node.