680 results about "State model" patented technology

A method and apparatus for reducing the workload of titrating drug to effect while leaving clinician users in control of a related procedure is described. A drug delivery device is controlled to achieve a target drug concentration at a selected site in the patient or a predetermined infusion rate waveform. The time profile of the target drug concentration or a predetermined infusion rate waveform is controlled by a drug state model that uses clinical heuristics to implement safe, pre-defined changes in the target drug concentration or infusion rate and user-commanded changes in target drug concentration or infusion rate. The invention allows time to assess the response of the patient to changes in drug level by making small incremental and conservative changes in drug level over time.

A method is disclosed for writing servo patterns for tracks on a rotating magnetic disk medium of a disk drive. Servo patterns are written on a reference track of the disk medium. Track following of the written servo patterns on the reference track is performed using a servo loop having a closed-loop response. A position error signal is generated for the reference track based on the track following. A correction signal is generated based on the track following using an observer of a one-dimensional state model that is equivalent to a two-dimensional state model of the servo loop. Servo patterns are written on a target track during track following of the reference track. The position error signal of the servo loop is adjusted based on the correction signal to reduce radial error propagation from the reference track to the target track.

A method is disclosed for writing servo patterns for tracks on a rotating magnetic disk medium of a magnetic disk drive. In the method, track following is performed along a reference track, defined by previously written servo burst patterns, using a servo loop having a closed-loop response. A position error signal is generated for the reference track based on the track following. A correction signal is generated based on the track following using an observer of a one-dimensional state model that is equivalent to the two-dimensional state model of the servo loop. Servo patterns are written at a target radial location on the magnetic disk medium during track following of the reference track. The position error signal of the servo loop is adjusted based on the correction signal to reduce radial error propagation from the reference track to the servo burst patterns at the target radial location.

Methods and arrangements for securing a software defined network. One or more control messages are intercepted. Information is extracted from the one or more control messages to create a global network state model, and there is determined, from the extracted information, presence of at least one prospective modification to the global network state model. Thereupon, a determination is made as to whether the at least one prospective modification presents a threat to security of the software defined network. Other variants and embodiments are broadly contemplated herein.

A trail management system for a transport network comprises a database storing data describing each of a plurality of actual and intended / planned trails within a network. Each trail is described in terms of a state model, in which the trail is recorded as being in one of a plurality of stable states, e.g. ready for service, not ready for service, existing in the network but not fully recorded in the network management system, or assigned for deletion from the network. The trail data is modified by a set of automatic processes, by means of which a trail data may transition between stable states. Each trail data may adopt one of a plurality of sub-states, e.g. fully provisioned, incomplete, in conflict with another trail, inconsistent with another trail, unprovisioned, or unsupported by network resources. The network management system learns of actual existing trails within the network and checks for consistency of actual trails with stored trail data.

The invention provides an intelligent control method and system for interactive home appliances on the basis of time-of-use electricity price response. The method is characterized by comprising the steps that the home appliances are automatically classified according to types of home appliance loads; use habits of the home appliances of users are determined according to historical home appliance running status data collected in 30 days; according to time-of-use electricity price policy information of a receiving power grid, basic load demand ranges corresponding to all time periods are predicted and determined for the users; according to obtained user home appliance use habit characteristics and load settings, controllable degree indexes of the home appliances are calculated, and running status models of the home appliances are determined; according to the controllable degree indexes of the home appliances, dynamic control priorities and a control algorithm of the home appliances are determined, and intelligent control over the home appliances is achieved. The method and system have the advantages that interactivity is good, the receiving degree of the users is high, the users can be guided to reasonably avoid the peak hours for electricity utilization, and electricity utilization cost is remarkably reduced.

System and method for parameterizing one or more steady-state models each having a plurality of model parameters for mapping model input to model output through a stored representation of an in-situ hydrocarbon reservoir. For each model, training data representing operation of the reservoir is provided including input values and target output values. A next input value(s) and next target output value are received from the training data. The model is parameterized with the input value(s) and target output value, and derivative constraints imposed to constrain relationships between the input value(s) and a resulting model output value, using an optimizer to perform constrained optimization on the parameters to satisfy an objective function subject to the derivative constraints. The receiving and parameterizing are performed iteratively, generating a parameterized model. Multiple models form an aggregate model of the system / process, which may be optimized to satisfy a second objective function subject to operational constraints.

The invention relates to a battery degradation state model-based lithium ion battery cycle life prediction method, aiming at solving the problem of difficulty in modeling existing in the conventional lithium ion battery cycle life prediction process. The battery degradation state model-based lithium ion battery cycle life prediction method comprises the following steps: 1, acquiring battery monitoring data and preprocessing the data; 2, obtaining a battery degradation state model according to battery degradation state model training; and 3, predicting the lithium ion battery cycle life according to the battery degradation state model obtained by the step 2 to obtain a lithium ion battery cycle life value so as to realize battery degradation state model-based lithium ion battery cycle life prediction. The battery degradation state model-based lithium ion battery cycle life prediction method is suitable for the field of batteries.

The city road net traffic flow state estimating method is based on vehicular GPS data combined with SCATS provided traffic signal state information. The present invention establishes city road net traffic flow state model by adopting unit oriented road sections as targets and through least-square method fitting in the three-dimensional space of distance, time and speed. Through the conversion from curved surface in the three-dimensional space to the curve in the two-dimensional space, the average speeds in various oriented road sections of the city road net in fixed time are obtained, and the current traffic flow jamming state is analyzed and estimated with speed as index. The traffic flow state estimation is completed through establishing the traffic flow state model for each of the unit oriented road sections between two signal lamps and connecting the curve surface models of adjacent equidirectional road sections.

The invention relates to a word cognitive state model-based adaptive learning system and method. The system comprises a mobile terminal, a computer terminal and a server end; the system further comprises a domain knowledge and extension resource module, a cognitive objective introduction module, a cognitive state analysis module, an adaptive learning module and a background database. The method comprises the steps of domain knowledge and extension resource structuralization, cognitive objective introduction and learning activity design, word learning and breakthrough making, learner cognitive state analysis and adaptive content recommendation. The word cognitive state model-based adaptive learning system and method of the invention are applicable to students who can learn English, can analyze the interactive data of the learners, evaluate the cognitive states of the learners, provide accurate recommendations which are suitable for the cognitive states and capacities of the learners in word learning, and minimize the cognitive burden brought by vocabularies which have been already mastered by the learners or vocabularies which are excessively difficult so as to reduce learning burden and improve learning efficiency.

The invention is applicable in the field of communications and provides a method and a device for tire pressure monitoring. The method comprises the following steps: tire pressure monitoring data is collected in a current state; the state model is switched according to the collected tire pressure monitoring data. According to the method and the device provided by the invention, the validity, reliability and accuracy of the tire pressure monitoring are guaranteed; simultaneously, the power dissipation is effectively reduced; the service life of a battery is prolonged.

The invention relates to an intelligent traffic lamp control method based on deep intensive learning. The method comprises the steps of selecting a center intersection, wherein multiple peripheral intersections communicated with the center intersection are arranged at the periphery of the center intersection; obtaining road condition information and signal lamp information of each intersection; building an intersection congestion and unblocking state model; modelling a traffic signal lamp control problem into a Markov decision-making process, and defining the state, motion and an immediate awarding function in the process; building a return value function model, utilizing a DQN deep intensive learning algorithm for solving an optimum strategy, and utilizing the optimum strategy for controlling traffic lights of all the intersections. By means of the method, a control strategy of the traffic lights can be self-adaptively and dynamically adjusted according to the real-time road conditioninformation. In the meanwhile, multiple intersections are synchronously adjusted, and full play can be given to the traffic capability of all the intersections.

There is provided an abnormality determination unit that determines whether or not an output measurement unit is abnormal, in order to determine abnormality of the output measurement unit. When the output measurement unit is abnormal, the operation of an elastic actuator is controlled not in accordance with a measurement result of the output measurement unit but in accordance with an internal state model.

The identification of Hammerstein models relates to a computerized method for identifying Hammerstein models in which the linear dynamic part is modeled by a space-state model and the static nonlinear part is modeled using a radial basis function neural network (RBFNN), and in which a particle swarm optimization (PSO) algorithm is used to estimate the neural network parameters and a numerical algorithm for subspace state-space system identification (N4SID) is used for estimation of parameters of the linear part.

The invention discloses a residual error posterior-based abnormal value online detection and confidence degree assessment method. The method comprises the steps of collecting data, establishing time series data, performing linear fitting on the time series data to obtain a linear combination formula of data at a current moment and p pieces of previous data, and predicting a data value of subsequent time; comparing the predicted data value with an actually detected data value to obtain a predicted residual error series; determining a probability density function of the predicted residual error series by adopting a KDE (Kernel Density Estimation) method; performing posterior ratio check on the predicted residual error series, and judging whether the data at the current moment is an abnormal point or not; and by taking the time series data as an input, building an SOM state model, obtaining state series and state transition probability matrixes, defining an abnormal scoring function, and outputting an abnormal score. By comparing the probability that the data is the abnormal point with the probability that the data is a normal point, the abnormal value in the pollutant discharge concentration time series data is identified online, so that the accuracy and reliability of abnormal value judgment are improved.

The invention belongs to navigation methods and particularly relates to a method for combined navigation of an inertia / visual odometer / laser radar. The method comprises (1) state model establishment, (2) visual odometer speed measurement based on characteristic information, (3) establishment of a measurement equation and obtaining of a measurement value, (4) Kalman filtering and (5) system error correction. The method has the advantages that a machine vision autonomous navigation technology is used, a monocular camera can measure a vector speed under the conditions of a known distance through the difference of a front frame image and a rear frame image, the laser radar can accurately measure the distance of an observation point and then measure a vector speed, navigation is performed by utilizing combination of the speed obtained through measurement and an inertial reference speed, and high-accuracy navigation is performed finally under the conditions of no outside reference information.

The invention provides a method for planning the routes of multi-unmanned aerial vehicles based on a particle swarm optimization algorithm. The method comprises the following steps: establishing a three-dimensional map for planning space of the routes of the multi-unmanned aerial vehicles at first; then constructing multi-unmanned aerial vehicle route planning models under the three-dimensional map, wherein the models mainly comprise a barrier model, a route model, an unmanned aerial vehicle state model, a constraint model and a multi-unmanned aerial vehicle route planning mathematic model; and solving the problem of multi-unmanned aerial vehicle route planning under the three-dimensional map by using the particle swarm optimization algorithm. The method provided by the invention improves multi-unmanned aerial vehicle route planning capacity in a complex environment and provides technical support for air traffic management platforms for unmanned aerial vehicles, autonomous flight systems for multi-unmanned aerial vehicles, etc.

The invention discloses an intelligent monitoring system and method for health of empty-nest aged people. The intelligent monitoring system comprises a multi-mode real-time detection module, a big data analysis module and a 4G network response module, wherein the multi-mode real-time detection module is used for collecting multi-mode health data of the empty-nest aged people and judging an urgent abnormal event; the big data analysis module is used for analyzing the health data, establishing a health state model and pre-warning an abnormal state; and the 4G network response module is used for pushing the urgent abnormal event and the abnormal state in real time and realizing interactive operation with a user operation terminal. Meanwhile, the intelligent monitoring system adopts a dual-warning mechanism, so that accurate and comprehensive intelligent monitoring of the physical state and mental state of the empty-nest aged people can be realized; and by utilizing the operation terminal, a guardian and a medical worker can timely obtain corresponding warning information and selectively check the health state of the empty-nest aged people at any time, and the monitoring manner is changed from a traditional passive accepting form into an automatic query form.

In a method for monitoring hardware devices of the hardware devices network to detect failures, past logged jobs are analyzed to determine a pattern of past usage of hardware devices by users of the hardware devices network. A failure of a hardware device is inferred based on a deviation in usage of the hardware devices from the pattern of past usage. In some embodiments, a state model is constructed that assigns one of two or more available operational states to each hardware device of the hardware devices network based on a recent load of jobs logged to the hardware device and a stored history of logged jobs indicative of typical loading of the hardware devices. One or more failing hardware devices are identified based on the assigned operational states.

The invention relates to an independent celestial navigation method for a Mars capturing section of a deep space probe. The method comprises the following steps of: establishing a state model of the deep space probe according to a circular restrictive four-body track dynamics model; acquiring picture element and picture line information of Mars, Martian satellites and a background fixed star, converting the acquired picture element and picture line information into angle information of the Mars, the Martian satellites and the background fixed star and establishing an angle measuring model of the Mars, a Martian satellite I and a Martian satellite II; and estimating the attitude information of the probe by adopting a q-method and estimating the position and speed of the deep space probe in combination with Unscented Kallman filtering. The method has high estimation accuracy, and is very suitable for independent navigation of Mars capturing. The method belongs to the technical field of aerospace navigation. By adopting the method, high-accuracy navigation parameters can be provided for Mars capturing of the deep space probe, and references can be provided for the design of an independent navigation system of the deep space probe.

The invention belongs to the technical field of mineral resources, and particularly relates to a method for computing unsteady state output of a shale gas reservoir complex fracture network. The method comprises the following steps that S1, basic parameters of shale gas reservoir strata and shale gas are collected, and a shale substrate porous media capillary tube permeability model is established; S2, shale gas reservoir microseism fracture imaging data is collected, a box method is applied to acquire fractal dimensions, a two-dimensional discrete fracture network model is established, and animproved cubic law is adopted to calculate the complex fracture network permeability; S3, it is considered that a shale gas reservoir fracturing horizontal well outer region is a substrate area infiltration area and a shale gas reservoir fracturing horizontal well inner region is a complex fracture network area flow area, and a shale gas reservoir fracturing horizontal well steady-state productivity model is established; according to a shale gas pressure propagation radius equation and a shale gas reservoir material balance equation, a continuous quasi-steady method is applied to establish ashale gas reservoir fracturing horizontal well unsteady state model. The method for the computing unsteady state output of the shale gas reservoir complex fracture network has the advantages that theconsideration factors are close to the actual conditions of strata, and computed results are more accurate.

The invention discloses a power distribution system probability reliability assessing method based on an analytical method. The method comprises the steps that a given power distribution network is partitioned into a fault area, an isolation area, a gapless island area and an influence-free area are classified, a fault mode impact analysis sheet base is established, and parameters are initialized; a stimulation clock is initialized, a random number is generated, the minimum non-failure operation time is obtained according to failure rate parameters of all element state models, the fault isolation time and the load band transferring time are worked out, and the stimulation clock is pushed; the fault mode impact analysis sheet base is inquired, classifications of all cells are determined, whether an island is formed or not is judged, and different methods are adopted for processing the island area and the non-island area; the state sample of an energy storage device is established according to the energy storage device electrical charge state probability distribution obtained through the probability reliability calculating method; distribution probabilities of single-time fault indexes of load points are overlapped, and the probability reliability indexes of the loads and the system are calculated. The simulation speed is increased while certain calculation precision is ensured, and the power grid situation is comprehensively reflected.

The amount of data being delivered across networks is constantly increasing. This system and method demonstrates an improved system and method for establishing secure network connections with increased scalability and reduced latency. This approach also includes arbitrary segmentation of incoming network traffic, and dynamic assignment of parallel processing resources to execute application code specific to the segmented packets. The method uses a modified network state model to optimize the delivery of information and compensate for overall network latencies by eliminating excessive messaging. Network data is application generated, and encoded into pixel values in a shared framebuffer using many processors in parallel. These pixel values are transported over existing high speed video links to the Advanced Network Interface Card, where the network data is extracted and placed directly on to high speed network links.

There is provided a system and a computer-implemented method of detecting malware in real time in a live environment. The method comprises: monitoring one or more operations of at least one program concurrently running in the live environment, building at least one stateful model in accordance with the one or more operations, analyzing the at least one stateful model to identify one or more behaviors, and determining the presence of malware based on the identified one or more behaviors.

The invention relates to an inertia / vision integrated navigation method for a deep-space detection patrolling device. The method comprises the steps that a state model of inertia / vision integrated navigation of the patrolling device is established according to a serial inertial navigation mechanical arrangement in a fixed coordinate system of a planet; a binocular vision camera is used for shooting a surrounding environment of the patrolling device; a stereoimage sequence is obtained; pixel coordinates of characteristic points are obtained by an image characteristic extraction algorithm; a measurement model by taking the pixel coordinates of the characteristic points as measurement vectors is established; and a position, a speed and a state of the deep-space detection patrolling device are estimated by using Unscented kalman filter. The method can effectively correct a position error of inertial navigation and improve the navigation accuracy, is very applicable to autonomous navigation of the deep-space detection patrolling device, belongs to the technical field of space navigation, can provide high-precision navigational parameters for the deep-space detection patrolling device, and can provide references for the design of an autonomous navigation system of the deep-space detection patrolling device.

The invention provides a method and a device for achieving safety testing on protocol implementation. The method includes: mopping protocol specifications into a first protocol state model; leading illegal state into the first protocol state model to form a second protocol state model; representing testing aims of the first protocol state model / the second protocol state model through assertions, and representing grammar rules of the messages input legally through grammar constrains; performing behavior mutation operations on the second protocol state model to generate a variant protocol state model and performing model detection, and using the obtained counter example mapping violating the assertions as a testing sequence; performing grammar mutation operations on the testing sequence to generate the testing sequence containing the variant testing messages, and using the testing sequence containing effective variant messages as a testing case; and using the testing case to perform muzzy testing on protocol implementation to detect the security holes in the protocol implementation. By means of an embodiment of the method and device for achieving safety testing on the protocol implementation, safety testing of the protocol can be achieved efficiently with low cost.

The invention provides a computing unloading scheduling method based on depth reinforcement learning, which provides a method for computing unloading to make unloading decision for Internet of Thingsequipment, including making decisions on various aspects needing to be unloaded according to the basic model of computing unloading. Based on different optimization objectives, different optimizationobjectives can be achieved by changing the value function. TheDeep-SARSA algorithm is similar to DQN algorithm, which combines reinforcement learning and depth learning. It can effectively change theunloading state and unloading action into training samples of depth learning when cooperating with experience pool. The invention can effectively carry out machine learning on an unloading state modelof an unlimited dimension, reducing the complexity of learning, this method uses neural network as the linear approximator of Q value, which can effectively improve the training speed and reduce thesample required for training. This method can effectively make the best decision through deep reinforcement learning under the given model and optimization objective.

The invention provides an intelligent battery state of health (SOH) detection method used for an electric vehicle sharing charging pile. The intelligent battery SOH detection method comprises steps that power battery data of user application is tracked, and the power battery data is categorized and stored; the actual measurement current, the end voltage, and the surface temperature of the power battery data are used as input samples, and the training data set of the power battery data is established, and then an LSSVM regression model is acquired; a power battery SOH training model is established according to parameters, and is used for health estimation of a current charging battery unit; a battery current maximum capacity is used as a SOH evaluation index, and a power battery SOH model is established, and is optimized; voltage, current, temperature, and time information of a database are transmitted to a monitoring platform by a communication protocol, and the monitoring platform is used to estimate the SOH of the battery. By adopting a big data analysis method, the establishing of the battery data model is completed, and during a charging process, SOH of energy storage systems of all of users is evaluated.