162 results about "Reasoning algorithm" patented technology

The technology of the 4D-GIS system deploys a GIS-based algorithm used to determine the location of a moving target through registering the terrain image obtained from a Moving Target Indication (MTI) sensor or small Unmanned Aerial Vehicle (UAV) camera with the digital map from GIS. For motion prediction the target state is estimated using an Extended Kalman Filter (EKF). In order to enhance the prediction of the moving target's trajectory a fuzzy logic reasoning algorithm is used to estimate the destination of a moving target through synthesizing data from GIS, target statistics, tactics and other past experience derived information, such as, likely moving direction of targets in correlation with the nature of the terrain and surmised mission.

A local-neighborhood Laplacian Eigenmap (LNLE) algorithm is provided for methods and systems for semi-supervised learning on manifolds of data points in a high-dimensional space. In one embodiment, an LNLE based method includes building an adjacency graph over a dataset of labelled and unlabelled points. The adjacency graph is then used for finding a set of local neighbors with respect to an unlabelled data point to be classified. An eigen decomposition of the local subgraph provides a smooth function over the subgraph. The smooth function can be evaluated and based on the function evaluation the unclassified data point can be labelled. In one embodiment, a transductive inference (TI) algorithmic approach is provided. In another embodiment, a semi-supervised inductive inference (SSII) algorithmic approach is provided for classification of subsequent data points. A confidence determination can be provided based on a number of labeled data points within the local neighborhood. Experimental results comparing LNLE and simple LE approaches are presented.

A system and method that translates sentences of natural language text into sets of axioms of formal logic that are consistent with parses resulting from NLP and acquired constraints as they accumulate. The system and method further present these axioms so as to facilitate further disambiguation of such sentences and produces axioms of formal logic suitable for processing by automated reasoning technologies, such as first-order or description logic suitable for processing by various reasoning algorithms, such as logic programs, inference engines, theorem provers, and rule-based systems.

A method and system for managing a policy includes, in response to determining the presence of a conflict, determining a semantic equivalence between a component of a policy rule and at least one additional policy rule. The determining a semantic equivalence is performed by using a semantic reasoning algorithm that includes the steps of determining a first policy target of a first policy rule and a second policy target of a second policy rule, determining a meaning of the first policy target and a meaning of the second policy rule, assigning a confidence value based on the determined meaning of the first policy, assigning a confidence value based on the determined meaning of the second policy, performing a semantic comparison between the first policy target and the second policy target, and determining, based at least in part on the semantic comparison, the presence of a conflict between the first and second policy targets.

The invention discloses a gas path fault diagnostic method for a marine gas turbine based on a fuzzy Petri net (FPN), comprising the following steps of: (1) obtaining the operating standard value of the marine gas turbine; (2) based on an actually-measured value, calculating the membership grade of the actual operating parameter deviation belonging to the criterion deviation fuzzy subset by using a membership function; (3) defining the FPN in response to the characteristics of the fault diagnostic Petri net after determining the operating standard value and the deviation; and (4) determining a FPN reasoning algorithm for gas path fault diagnosis for the marine gas turbine based on related definitions of the transitional enabling and igniting rules. Through the gas path fault diagnostic method for the marine gas turbine based on the fuzzy Petri net (FPN), the defects of not strong university, wrong judgment and reasoning of the traditional method are overcome; it is actually proved that the problems of fuzziness and nondeterminacy in an gas path fault can be effectively solved by the gas path fault diagnostic algorithm for the marine gas turbine established by the gas path fault diagnostic method disclosed by the invention.

The invention relates to a multi-person attitude estimation method based on a multi-layer fractal network and a joint relative pattern. The multi-person attitude estimation method adopts a three-layerfractal network model to predict key points of a human body, proposes a hierarchical bidirectional reasoning algorithm to match joint points of a plurality of people, realizes the optimal matching ofa plurality of human body joint points according to a relative degree and external space constraint relationship between each pair of joint points, effectively removing disorder matching among a large number of joint points, and can greatly improve the average precision of multi-person attitude estimation.

The invention discloses a meteorological threat assessment method based on a discrete dynamic Bayesian network. The method comprises the following steps: collecting an observed weather type, intensity information and UAV (Unmanned Aerial Vehicle) position and attitude information; performing a quantization treatment according to a divided quantization level, and establishing an observation evidence list; using expert knowledge or experience to establish a conditional probability transfer matrix between states, and determining a state transfer matrix between time slices; establishing a discrete dynamic Bayesian network model between a meteorological threat level, a meteorological factor and the UAV; and using a Hidden Markov Model reasoning algorithm to calculate the final meteorological threat level. The meteorological threat assessment method based on the discrete dynamic Bayesian network provided by the invention realizes the organic combination of a continuous observation value and the discrete dynamic Bayesian network, and reasons out the probability distribution of a meteorological threat degree in combination with the HMM (Hidden Markov Model) reasoning algorithm, so that the effectiveness, the practicability and the accuracy of meteorological assessment can be greatly improved.

The invention relates to a multi-scale region flood disaster risk remote sensing assessment method and system. The method comprises the steps of cooperatively extracting feature information of a flood disaster water body by utilizing multi-source multi-platform remote sensing data to form space distribution data of the water body; by analyzing critical features of the flood disaster water body, and in combination with environmental factors in a historical flood disaster occurrence and development process, establishing a flood disaster long-time series data set based on the critical features of the water body; and then according to a reasoning algorithm, performing multi-scale region flood disaster risk level assessment from a disaster flooding range and a disaster loss degree to generate a space distribution result of flood disaster risk assessment. The method has a very important effect for enhancing disaster early warning, avoiding risk and the like, can reduce the loss caused by flood disasters maximally, and is an important scientific basis of performing disaster risk management and decision-making.

A method, apparatus and computer program product for conversion of decision trees into probabilistic models such as Bayesian networks. Decisions trees are converted into probabilistic models without loss of information stored in the decision tree implicitly or explicitly. As a result, the probabilistic model is usable to reproduce the paths of the original tree. An inference algorithm can be used to reproduce the paths of the original tree from the probabilistic model.

The invention discloses a knowledge map relation inference algorithm based on a stacked neural network, and belongs to the technical field of artificial intelligence representation learning. The invention constructs a stacked neural network model comprising two components: a standard LSTM loop neural network and a multi-layer perceptron network. The model regards the triplets in the knowledge mapas short sentences and uses the learning ability of LSTM loop neural network to model the logical and semantic characteristics of the knowledge map in order to learn the grammatical and semantic information in the knowledge map. Through the bottom-level feature learning process, the upper-level fully connected network can provide enough discrimination to distinguish objects in different contexts.The algorithm of the invention models the three tuples in the knowledge map from the perspective of semantics, fundamentally solves the semantic diversity of entities and relationships, can greatly reduce the computational overhead of the relational reasoning of the large-scale knowledge map, and simultaneously can ensure the effectiveness of the relational reasoning.

The invention discloses a two-dimensional video depth map generation process, which includes: (1), acquiring two-dimensional video and marking the key frames and non-key frames; (2) segmenting images and performing artificial depth assignment to the key frames; (3) computing depth information of the non-key frames according to depth information of the key frames; and (4), recovering a non-key frame depth map to synthesize a depth map of the whole video. The key frame images are subjected to artificial depth assignment, other images are computed by the depth map reasoning algorithm on the basis of region segmentation, and accordingly labor cost for making the depth map of the whole video is reduced greatly, the corresponding speed and efficiency is improved, the acquired depth map accords with the real depth information, and quality and effect of the video depth map are guaranteed effectively. The two-dimensional video depth map generation process can be implemented on hardware systems such as general PC (personal computers) or work stations and the like.

The invention discloses a human body detection and attitude estimation combined deep network learning method. The method comprises: combining a CNN model with overall and local information to carry out detection, the model efficiently extracting underlying features through a shared convolution layer, and then the features respectively passing through two branches connected in parallel to carry outhuman body detection and attitude estimation. According to the invention, a Fusion model with a hidden tree structure reasoning algorithm is used to fuse results of human body detection and attitudeestimation, so that a robust and reliable human body detection box is obtained. According to the method, through an NMS algorithm (poseNMS), the obtained information of the human body parts is utilized, and all the individuals which are shielded with one another are effectively reserved. According to the method, a tree structure model is used for embedding information of each part into a detectedbounding box, and a convolutional network is used for realizing an inference algorithm. The method integrates the advantages of overall modeling and local modeling, has a very good detection effect oncrowded and shielded people and pedestrians with uncommon posture behaviors, and can be better integrated into practical application.

The invention discloses a fast multi-channel combined detecting cognizing method based on reasoning in a wireless communication system and is mainly characterized in that each secondary user in the wireless communication system randomly selects partial channels in the channels to be cognized according to an ideal arc wave distribution in each cognizing process for carrying out cognizing, and the channel marks of the partial channels and the cognizing result of the partial channels are sent to a fusing center; then the fusing center utilizes the reasoning algorithms like belief-propagation to carry out comprehensive analyzing and judging on the moving state of the main users on all the channels. The method can quickly, effectively and reliably detect the moving states of the main users on a plurality of channels and can reduce the sensing capacity demand of each secondary user for adapting different network scales, thus greatly improving the detecting efficiency, reducing the time delay of sensing, and effectively improving the operating factor of frequency spectrum.

The invention discloses an unmanned ship used for water quality monitoring and pollution source tracking and a pollution source tracking method. The unmanned ship comprises a catamaran composed of twoclosed cabins and a connecting plate; a solar cell panel for charging lithium batteries is laid on the upper surface of the catamaran; a U-shaped equipment rack is arranged above the catamaran; a video camera, a radar and an antenna are fixedly arranged on the equipment rack; and a control cabin, a battery cabin, a generator cabin, a water quality sampling cabin and five equipment cabins for placing water quality monitoring instruments are arranged inside each of the two closed cabins respectively. The unmanned ship disclosed by the invention has the functions of tracking and tracing water pollution sources, the concentration field of pollutants inside water is detected through a full spectrum water quality analyzer due to the tracking and tracing functions, in combination with the turbulent model of the high Reynolds coefficient of a water flow, a Bayesian reasoning algorithm with the optimal vector of phi is adopted, and the unmanned ship is automatically controlled to cruise to thesource of the pollution source, so that the tracking and tracing of water pollution are realized.

The invention discloses a power system fault automatic diagnosis method based on a fuzzy reasoning spiking neural P system. Firstly, topological data representing the whole topological structure of apower network, and protection data of protection configuration of a power grid network, are constructed. Secondly, after a SCADA system provides action information of protection and switching, a powernetwork topology analysis algorithm program is called to determine related suspicious fault elements based on on-off states of a breaker, a fuzzy reasoning spiking neural P system diagnosis model corresponding to each suspicious fault element is generated, and then, a reasoning algorithm of the fuzzy reasoning spiking neural P system is called to determine the fault elements. In the method provided by the invention, fault diagnosis of the fuzzy reasoning spiking neural P system in a power transmission network is realized through a programmed mode. The whole fault diagnosis process is realizedautomatically, even under a situation of complex faults of a large-scale power network and a situation where the SCADA system provides complete fault information or information, such as information of maloperation or failure to operate of a protection breaker, is not complete, correct diagnosis results can be obtained efficiently and automatically.

The invention relates to a lattice implication reasoning algorithm of a bug in a partitioning protection system of a smart grid, comprising the steps of: building a lattice implication algebraic system model according to the experimental environment of the partitioning protection system of the smart grid; describing the bug in the model by a method of a concept lattice; building an automatic reasoning system of the lattice implication on the basis; reasoning and inspecting the bug of the system; and inspecting the reasoning efficiency by the experiment environment of the system. The automatic reasoning system of the boundary bug is built in the partitioning protection system of the smart grid by adopting a method that lattice implication is combined with concept lattice, so that the bug detection efficiency of the smart grid is improved.

The invention discloses a power supply reliability-oriented secondary optimization evaluation model for the reliability of a power distribution network. The evaluation model comprises a smallest isolation area sub-model, a failing area influence sub-model and a load area statistical sub-model, wherein a double-layer heterogeneous structure is adopted for each of the sub-models and the evaluation model, and comprises a probability layer and a frequency layer, which are used for calculating a power failure probability index and a power failure frequency index respectively, graphic structures of the probability layers and the frequency layers are the same, and a layer of graphic structure can correspond to two layers of data during model storage. According to the power supply reliability-oriented secondary optimization evaluation model for the reliability of the power distribution network, an intuitive graphical method, a solid probability theory basis and an effective reasoning algorithm are used for performing modeling and reliability analysis on the power distribution network.

The invention discloses a real-time warning method under a space-time dynamic mode of a sensor network. The real-time warning method comprises the following steps: using a real-time database system to acquire real-time dynamic data from the sensor network in a real-time manner in advance; defining atomic events based on an ECA (Event Condition Action) rule in advance, and compounding the atomic events to obtain a complex event; establishing a dynamic Bayes network model in advance to obtain an indeterminate complex event rule base; establishing a detection method for carrying out real-time analysis and matching on event flows based on an automatic mechanism in advance; establishing a reasoning structure based on an intelligent agent dependency graph in advance; reasoning the event flows from the sensor network on the reasoning structure based on the intelligent agent dependency graph in a real-time reasoning algorithm based on the ECA rule to obtain a determined result of the events; and trigging the real-time warning when the determined result is true, actuating corresponding inform operation, and warning. The real-time warning method has the advantages of real-time performance, stable property and higher efficiency.

The invention discloses a personal data analysis method based on a Bayesian network and a computer storage medium, and the method comprises the following steps: (1) enabling personal life behavior data to be embodied as a one-dimensional vector of behaviors and behavior attributes, enabling the behavior attributes to at least comprise a time attribute, and obtaining a life behavior data record through data preprocessing; (2) learning the data through a hybrid structure learning algorithm, and constructing a life data Bayesian network; (3) parameter learning is carried out according to the lifedata Bayesian network, and a conditional probability distribution table of each network node is obtained through learning; and (4) calculating the probability of occurrence of other behaviors based on the probability of the specific behavior by using a joint tree reasoning algorithm according to the life data Bayesian network, and completing the analysis and prediction of the personal life behavior. According to the method, the Bayesian network is applied to personal behavior data analysis, and the network construction method is improved, so that the learning accuracy and the convergence of the algorithm are effectively improved, and the operation performance is improved.

The invention provides a man-machine conversation method, a man-machine conversation device, a storage medium and a computer program product. The method comprises the steps of determining a current conversation theme and current utterance information of a user; determining a current utterance representation vector of the user according to the current utterance information; in combination with thecurrent utterance information and the current utterance representation vector, performing graph reasoning calculation on the heterogeneous knowledge graph corresponding to the current dialogue theme,and selecting current knowledge corresponding to the current utterance information from the heterogeneous knowledge graph; acquiring current utterance information according to the current utterance information and current knowledge; generating a reply statement corresponding to the current statement, wherein the heterogeneous knowledge graph is created on the basis of structured knowledge and unstructured knowledge and can generate reply statements with rich contents. In addition, the accuracy of knowledge selection can be improved by adopting a graph reasoning algorithm, so that the knowledgeselection process has very good interpretability and generalization ability. Meanwhile, the dependence of the whole scheme on corpora with labels is reduced.

The invention provides a construction engineering project cluster establishment method introducing a cloud model for evaluation and selection. The construction engineering project cluster establishment method comprises the steps that step 1, a construction engineering project cluster establishment selection and evaluation index system is established; step 2, variable clouding is performed, and all parameters of the cloud model are confirmed according to change range of evaluation indexes and the definition interval of evaluation scores; step 3, the scores of all the evaluation indexes are calculated by utilizing a cloud uncertainty reasoning algorithm; step 4, weight coefficient of each evaluation index is confirmed according to an entropy weight method; step 5, priority assessment is performed on engineering projects to be selected; and step 6, the most reasonable construction engineering project cluster is established. The cloud model in the field of artificial intelligence is introduced to the construction engineering project cluster establishment selection and evaluation in view of fuzzy and random problems existing in the construction engineering project cluster establishment selection and evaluation process so that effective conversion between a qualitative concept and quantitative numerical representation thereof can be realized, and the evaluation result is more objective and accurate.

The invention discloses a knowledge-based programmable logic controller with flexible in-field knowledge management and analytics. A method of operating an intelligent programmable logic controller over a plurality of scan cycles includes the intelligent programmable logic controller executing a control program and one or more reasoning algorithms for analyzing data received and transmitted by the intelligent programmable logic controller. The intelligent programmable logic controller receives one or more user-specified declarative knowledge models from an external source via a deployment interface included in the intelligent programmable logic controller. The intelligent programmable logic controller dynamically modifies the reasoning algorithms during runtime of the control program based on the user-specified declarative knowledge models.

The invention belongs to the technical field of information extraction, and particularly relates to a knowledge verification model construction and analysis method based on probability soft logic, which comprises the following steps of: a, forming a candidate knowledge set by knowledge extracted from webpage web texts of a plurality of data sources by an information extraction system; b, carryingout credibility calculation on the candidate knowledge set; c, performing logic predicate representation on each entity in the candidate knowledge set; d, constructing a first-order logic rule of theknowledge verification model based on the entity analysis and the ontology constraint, generating the first-order logic rule in the probability soft logic model through the constructed logic rule, andachieving entity relationship and entity label verification in the candidate knowledge set; and e, setting probability distribution of the knowledge verification model, and calculating and selectingcorresponding knowledge to be updated through an inference algorithm. According to the method, the candidate knowledge set is verified, so that the accuracy of the candidate knowledge set is greatly improved.

The invention relates to a fault location method of a distributed generation including power distribution network of a synaptic plasticity based SNP system. The fault location method comprises the following steps of (1), selecting a power cut interval; (2), establishing a fault location model of the distributed generation including power distribution network of the synaptic plasticity based SNP system; (3), determining a fault section; according to fault information and a forward synaptic matrix which are read by the system, carrying out operation on the established fault location model; (4),verifying fault location accuracy; according to a fault location result obtained by a reasoning algorithm and the value of a sequential pulse train in a bidirectional functional neuron O, verifying the accuracy of the fault location result by combining an actual current matrix C and the reasoning algorithm; and (5), completing the fault location and the verification of the fault location accuracythrough a fault judgment standard and a fault current information verification standard. The fault location method of the distributed generation including power distribution network of the synaptic plasticity based SNP system, which is provided by the invention, has high accuracy and high reliability, and can be widely applied to the fault location of the distributed generation including power distribution network.

The invention discloses an industrial control system attack clue discovery system based on a knowledge graph. Most industrial control systems are designed and developed before many years, lack corresponding security consideration, and inevitably have a lot of vulnerabilities endangering system security, and these vulnerabilities are likely to be utilized by intruders. Aiming at the fact that an industrial control intrusion detection system can only discover attacks but cannot provide clues related to the attacks, and the clues play an important role in rapid recovery of the system after the attacks, the industrial control system vulnerability utilization knowledge graph is constructed, and the related clues of the attacks are given from the perspective of vulnerability utilization. In theprocess of constructing the knowledge graph, an attack information named entity identification method based on a conditional random field, an entity alignment framework based on rule and character similarity calculation and a knowledge reasoning algorithm based on type limitation and pre-training model negative triple potential correct probability are provided. According to the method, the knowledge graph is visually displayed in a force-oriented graph mode according to the attack clues obtained through user input, and the method is more accurate and visual.