49 results about "State transition function" patented technology

To execute all processing in an array section of an array-type processor, each processor must execute processing of different types, i.e., processing of an operating unit and processing of a random logic circuit, which limits its size and processing performance. A data path section including processors arranged in an array are connected via programmable switches to primarily execute processing of operation and a state transition controller configured to easily implement a state transition function to control state transitions are independently disposed. These sections are configured in customized structure for respective processing purposes to efficiently implement and achieve the processing of operation and the control operation.

A programmable finite state machine (FSM) includes, in part, first and second memories, and a selection circuit coupled to each of the memories. Upon receiving a (k+m)-bit word representative of the k-bit input symbol and the m-bit current state, the first memory supplies one ore more matching transition rules stored therein. The selection circuit selects the most specific of the supplied rules. The transition rules are stored in the first memory in a ranking order of generality. The second memory receives the selected transition rule and supplies the next state of the FSM. The first memory may be a ternary content addressable memory and the second memory may be a static random access memory. The contents of both the content addressable memory and the static random memory is determined by an algorithm which minimizes the number of terms required to represent the next-state transition functions.

A random binary sequence generator for generating a random binary sequence adapted to be used for producing random numbers, includes at least one logic circuit corresponding to an associated finite-state machine having a state-transition function including states arranged to form cycles of states, wherein the at least one logic circuit has a set of logic circuit inputs and a set of logic circuit outputs fed back to the logic circuit inputs; the associated finite-state machine is autonomous and asynchronous; the state-transition function is void of loops; and any of the cycles of states has either a minimum length equal to three states, in case the cycle is stable, or a minimum length of two states, in case the cycle is meta-stable.

The invention relates to a method of optimizing a state transition function specification for a state machine engine based on a probability distribution for the state transitions. For the preferred embodiment of the invention, a B-FSM state machine engine accesses a transition rule memory using a processor cache. The invention allows improving the cache hit rate by exploiting the probability distribution. The N transition rules that comprise a hash table entry will be loaded in a burst mode from the main memory, from which the N transition rules are transferred to the processor cache. Because the comparison of the actual state and input values against each of the transition rules can immediately start after each of these rules has been received, the overall performance is improved as the transition rule that is most likely to be selected is the first to be transferred as part of the burst access.

A programmable finite state machine (FSM) includes, in part, first and second memories, and a selection circuit coupled to each of the memories. Upon receiving a (k+m)-bit word representative of the k-bit input symbol and the m-bit current state, the first memory supplies one ore more matching transition rules stored therein. The selection circuit selects the most specific of the supplied rules. The transition rules are stored in the first memory in a ranking order of generality. The second memory receives the selected transition rule and supplies the next state of the FSM. The first memory may be a ternary content addressable memory and the second memory may be a static random access memory. The contents of both the content addressable memory and the static random memory is determined by an algorithm which minimizes the number of terms required to represent the next-state transition functions.

A simulation model conversion method based on a distributed interactive simulation platform, includes the steps: (1)determining the time management mechanism used in the simulation model conversion process to make the federal member of the simulation model unitedly order the insulated clock of the distributed interactive simulation platform, executing the simulation step length propulsion of the simulation model according to the simulation clock of the distributed interactive simulation platform; (2)dividing the intrinsic function of the simulation model to be converted into two parts: one part being a state transition function, the other part being an event handling function; (3) according to the state transition function and the event handling function, determining the model state parameters of the simulation model to be converted ordered and issued for the distributed interactive simulation platform and the special event ordered and issued for the distributed interactive simulation platform, forming the interface describing file; (4) according to the interface describing file, generating the object class library and the interactive class library for the conversion of the model to be converted; (5) according to the generated object class library and interactive class library, reforming the model to be converted and completing the conversion of the simulation model.

The invention discloses an extended finite state machine, and also discloses a method for simultaneously searching multiple regular expressions by using the extended finite state machine. The extended finite state machine adds four parameters, including the state identification set, the state transition function identification set, the mapping function between the state and the state identification, and the mapping function between the state transition function and the state transition function identification. The conversion function also introduces a flag, which allows users to implement efficient parallel search of multiple regular expressions without writing additional routines, preserves the information of a single regular expression, and effectively handles string fallback.

The invention discloses an FSA based voice emotion interaction device and method. The interaction device comprises a voice collection module, a voice emotion identification module and a voice emotion interaction module, wherein the voice collection module sets a record source, the sampling rate, an audio channel and the audio data format, and then writes raw data in an original file to generate a file A; the voice emotion identification module identifies the voice emotion of the file A; and the voice emotion interaction module makes emotion interaction via an emotion carrier. The voice emotion interaction module comprises an establishment module of an FSA model, an establishment module of an emotion interaction state transition table and a transition torque module, wherein the establishment module of the FSA model uses an FSA to describe voice emotion state and transition process thereof; the establishment module of the emotion interaction state transition table defines and determines transition function of the FSA; and the transition matrix module uses a state transition matrix to describe the state transition function in the FSA model.

The invention belongs to the technical field of intelligent robots, and concretely relates to an intrinsically motivated extreme learning machine autonomous development system and an operating method thereof. The autonomous development system comprises an inner state set, a motion set, a state transition function, an intrinsic motivation orientation function, a reward signal, a reinforced learning update iteration formula, an evaluation function and a motion selection probability. According to the invention, an intrinsic motivation signal is utilized to simulate an orientation cognitive mechanism of the interest of people in things so that a robot can finish relevant tasks voluntarily, thereby solving a problem that the robot is poor in self-learning. Furthermore, an extreme learning machine network is utilized to practice learning and store knowledge and experience so that the robot, if an experience fails, can use the stored knowledge and experience to keep exploring instead of learning from the beginning. In this way, the learning speed of the robot is increased, and a problem of low efficiency of reinforced learning for single-step learning is solved.

The invention discloses a method for the optimized dispatching of an extension type video flow in a partially observational Markovian decision process. The method simplifies an environment under radiobroadcasting, dispatches the extension type video flow under the condition of user state indetermination or partial observability, and establishes a data packet dispatching optimization model by thepartially observational Markovian decision process. The method comprises a state aggregate, a movement aggregate, a state transition function, a retribution function, an observation aggregate, an observation probability and a dispatching startup process. The method comprises the following steps: (1) assuming a radio broadcasting transmission environment model; (2) dividing each frame of data of the extension type video flow into layers, packaging each layer as a data packet, aggregating the data packet of each frame and establishing a data packet dispatching optimization model; and (3) optimizing and dispatching the extension type video flow. The method establishes the data packet dispatching optimization model of the extension type video flow, can enhance the average PSNR value of the video flow and realizes the optimization of the whole video reception quality of a user.

The invention discloses a random set theory-based multi-target tracking method. The method includes the following steps that: a multi-target state space is expanded, model dimension is increased based on original kinetic information, and therefore, characterization of target model information can be realized; a state transfer function and a likelihood function are expanded based on a jump Markov system, so that the state transfer function and the likelihood function can contain the model information; and the prediction and update process of an expanded multi-model generalized label multi-target Bernoulli filter is realized, and target states are extracted, and a target motion model is estimated. With the method adopted, problems in maneuvering multi-target tracking can be solved. The method has the advantages of high robustness, high adaptability and high estimation accuracy. With the method adopted, the problem of high multi-target maneuverability and inconsistency in practical application can be solved, and maneuvering multi-target tracking and target motion model estimation under complex scenes can be realized.

A closed loop current controller for an electromagnetic rail gun. It is necessary to control the muzzle velocity of a rail gun accurately for the gun to be a good artillery device. The present invention provides a closed loop control system to accurately regulate the energy transfer to a rail gun projectile and control its muzzle velocity. The rail gun control system includes a state space (state domain) control concept adapted to discrete control events that transition the system from one state to another until the final desired state (i.e., muzzle velocity) is reached. The control regulator preferably generates state transition functions that transition the projectile from state to state according to a defined current profile to provide a specified projectile muzzle velocity. The rail gun closed loop current controller also includes current reference compensation to correct for errors in previous state transitions.

The invention discloses a method for searching session on the basis of partially observable Markov decision process models. The method includes steps of 1), carrying out preparation phases, to be morespecific, marking corpora on the basis of themes used as units, recording interaction processes when the themes are retrieved by users and generating training files after all retrieval interaction isrecorded; 2), carrying out training phases, to be more specific, initializing statistics, computing initial state transition functions, parsing interactive data of the session from the training filesand updating the state transition functions according to the interactive data; 3), carrying out retrieval phases, to be more specific, receiving initial query statements, and updating the query statements when the users dissatisfy with fed documents until satisfactory document results are fed to the users.

The invention relates to a method and device for measuring regular expression state complexity. The method for the measuring regular expression state complexity comprises the steps that firstly, the curly relationship between any two states p and q in a given non-deterministic type finite automation M is judged, and the curly relationship is one of the five kinds of relationships: the mutual-exclusion relationship, the equivalence relationship, the included relationship, the including relationship and the independent relationship, M = (Q, sigma, delta, q0, F), the Q is a finite set, each element of the Q is called as a state, the sigma is a finite alphabet, each element of the sigma is called as an input character, the delta is a state transition function, and the q0 belongs to the Q, the q0 is the unique starting state, and the F is a terminal state set; two, the regular expression state complexity is estimated according to the judging result of the step one, the state complexity is the state number / Q' / of the deterministic type finite automation M'obtained by determining the M, and the M'=( Q', sigma, delta', q0', F'). The method and device for measuring regular expression state complexity can obtain the reasonable estimating value rapidly, and improve measuring efficiency.

The invention discloses a cognitive and learning method of a cognitive moving system with an inner engine. The cognitive model system to study the basis of automatic machines, including perceived state set, the set of actions, orientation map collection, curiosity, orientation function, learning orientation matrix, the state transition function and knowledge entropy, more than 10 parts. Perception of the current state of the system model; select action mechanism based on the engine; perform an action, state transition occurs; calculated value orientation function; Updates 'Perception - Movement'mapping; the process is repeated until the entropy minimal knowledge or learning time is greater than terminated time. The invention introduces the engine mechanism with active learning environment, not only the system has a strong self-learning and self-organizing ability, and can effectively avoid damaging small probability events, improve the stability of the system, for the establishment of having robot cognitive development capability to provide a strong foundation.

The invention discloses a system and a method for pet automatic feeding control on the basis of reinforcement learning. The system mainly comprises a base body and a controller, wherein the controllerobtains an optimal feeding strategy through the following method: A, establishing a pet feeding model; B, modeling a pet feeding problem into a Markov decision process model, and modeling a state, anaction, a state transfer function and an immediate reward function in the Markov decision process model; C, establishing a return value function model; and D, solving the optimal feeding strategy byutilizing a Sarsa reinforcement learning algorithm; and the controller controls the amount of feeding every time according to the optimal feeding strategy. The system and the method provided by the invention facilitate saving food, is applicable to eating habits of various pets, can acquire general feeding data of various pets by self-learning without the need of supplying, can obtain the optimalfeeding strategy which is more accurate than a conventional feeding manner, and can dynamically amend the optimal feeding strategy according to changes of the environment.

The invention discloses a large signal modeling method for a dual-active full-bridge converter under dual phase shift modulation. For a DAB converter for dual phase shift modulation, a precise discrete iterative model of the converter is established, then a state matrix and an input matrix of a large-signal equivalent circuit are obtained based on a state transfer function in a discrete domain byutilizing a Baker-Campbell-Hausdorff theorem, and a large-signal equivalent circuit model of the converter is constructed. According to the method, the defects of a traditional discrete iterative model are overcome, meanwhile, a mathematical model and a circuit model are unified, the coupling relation and the control-output transfer function of all variables of the system are visually reflected, and a new thought is provided for related research work such as control and stability analysis of the dual-active full-bridge (DAB) converter.

In one embodiment, method for controlling multiple wireless access nodes includes receiving, by a central controller from a base station (BS), a message indicating a channel state information (CSI) and determining a state transition function in accordance with the message. The method also includes determining a belief state in accordance with the state transition function and determining cooperation for a plurality of BSs including the BS in accordance with the belief state to produce a cooperation decision. Additionally, the method includes transmitting, by the central controller to the BS, the cooperation decision.

The invention discloses an inversion model updating method based on M-H sampling of Gaussian distribution and relates to the technical field of geophysical inversion. The method comprises the following steps of: step 1, pre-processing earthquake information to obtain an initial model of the parameters to be inverted; step 2, selecting the initial model of the t-th parameter to be inverted to calculate an initial wave impedance logarithm L1; step 3, calculating the variance of the initial wave impedance logarithm L1, establishing a state transition function obeying a Gaussian distribution, andrepeatedly updating the initial wave impedance logarithm L1 according to the state transition function; and step 4, determining whether t is greater than the seismic trace in the earthquake information; if so, ending the update; and if not, causing t=t+1 and jumping to step 2 to continue the inversion. The inversion model updating method based on M-H sampling of Gaussian distribution solves the problems of a too large convergence value, slow convergence speed, and local optimum due to the lack of the analysis of actual data when adopting M-H sampling based on uniform distribution in the priorinversion model updating, thereby minimizing the scope of convergence values, and being capable of quickly finding the optimum.

The invention relates to a description method of a virus propagation process considering the migration nodes of a dynamic network structure and belongs to the intersection field of complex network topology dynamics, biomathematics and public information security. The invention mainly aims to solve the problem that a current virus model fails to accurately describe the propagation process of the virus in a network. According to the method of the invention, reasonable settings and hypothesis descriptions for an S-to-I state transition function and a node migration function on the basis of a classical SIRS model, so that a virus propagation model comprehensively considering the migration nodes of the dynamic network structure is constructed.

The invention relates to a method of optimizing a state transition function specification for a state machine engine based on a probability distribution for the state transitions. For the preferred embodiment of the invention, a B-FSM state machine engine accesses a transition rule memory using a processor cache. The invention allows improving the cache hit rate by exploiting the probability distribution. The N transition rules that comprise a hash table entry will be loaded in a burst mode from the main memory, from which the N transition rules are transferred to the processor cache. Because the comparison of the actual state and input values against each of the transition rules can immediately start after each of these rules has been received, the overall performance is improved as the transition rule that is most likely to be selected is the first to be transferred as part of the burst access.

The invention discloses a power distribution network information physical element modeling method and system based on a finite state machine. The method comprises the steps: obtaining a power distribution network information physical element finite state set and a working function set in each state; according to an external interaction control mechanism between the elements, acquiring the state conversion rules of various elements under different external driving conditions; and constructing a finite state machine model of each type of element according to various attributes of the element, wherein the attributes of the element comprise a finite state set, a working function set in each state, a state conversion rule, a preset state transition function and a preset initial state set. The finite-state machine model constructed by the invention can sense events of the whole system, not only can trigger which control strategy to be adopted on a macroscopic level, but also can enable the state of the element to become the input of a control algorithm on a microscopic level, can control and optimize different levels of a power grid, and can better carry out scientific decision making and accurate execution.

The invention belongs to the field of operation control of a new energy power system, and particularly relates to a wind farm power uncertainty prediction method, which comprises the following steps of: dividing the wind speed into s wind speed sections from the cut-in wind speed to the cut-out wind speed according to a prediction accuracy requirement; calculating the probability density functionof the output power of the wind farm in any wind speed section; dividing arbitrary probability density function curve into np normal distribution curves by a P-Q segmentation method and a normal transformation method; fitting the normal distribution curve of arbitrary wind speed section by a cloud model, estimating a segmented cloud model parameter value by a Bayesian estimation method, and establishing a wind farm power uncertainty mapping model under arbitrary wind speed section; and predicting the wind farm power based on the state transfer function of a Markov model. By introducing the cloud model, the method can fully consider the uncertainty between the wind speed of the wind farm and the power, and significantly improve the power prediction accuracy of the wind farm.

Systems and methods are provided for efficiently counting detected events via a multiplicative group counter. An equivalent class polynomial congruent with a first of a plurality of elements comprising a multiplicative group is represented as a series of binary values. The represented polynomial is subjected to a state transition function as each event is detected, such that the series of binary values is altered to represent a new equivalent class polynomial congruent with a second of the plurality of elements of a multiplicative group. The series of binary values is decoded to determine a number of detected events recorded by the counter.

The invention discloses an improved algorithm for a trusted ant colony opportunity route in a family under an MSN. The method comprises the following steps: searching an optimal path in a network by utilizing distributed and adaptive characteristics of an ant colony algorithm, and evaluating a node comprehensive trust value by utilizing historical interaction information between nodes so as to effectively select trusted neighbor nodes in the network; then selecting relay nodes with higher forwarding willingness and tighter relationship for message transmission based on social intimacy state transition function, and finally realizing a credible message forwarding strategy in a family. The method has good performance in the MSN containing the selfish nodes, the selfish nodes can be effectively prevented from participating in the routing forwarding process, the messages are forwarded to the relay nodes with higher forwarding intention and tighter relationship, and the transmission reliability and the message forwarding performance of the messages are remarkably improved.

The invention provides an event processing method and an event processing system based on the risk level; a state automation machine is established by extracting the state of an event; the state automation machine processes the input event, and the method comprises steps of quickly judging whether the input event is in an event list or not, finding a corresponding state transfer function if the input event is in the event list, judging whether the event is a risky event or not through the state transfer function and feeding back to a client terminal; the system comprises a receiving module, an establishing module, an inquiry module and a judging module; the state of the event is extracted through the receiving module and the state automation machine is established by the establishing module; and the inquiry module determines whether the input event is in the event list in the station automation machine and carries out corresponding operation; the judging module determines whether a current file is a timeout file or not and processes correspondingly; and finally a processing module analyzes which events need to be sent to the client terminal.