29 results about "Recursive model" patented technology

The invention relates to a model for evaluating and predicting mild cognitive impairment risk of old people in a pension institution. The model sequentially comprises a behavior analysis module, a language recognition module, a social modeling module, a toughness calculation module, a feature fusion module, a risk reasoning module and the like. Behavior deviation characteristics and abnormal time periods are extracted by collecting behavior data of daily life, diet, social contact and the like of old people and comparing the behavior data with an institution work and rest template; in combination with nursing records, extracting language anomaly features; analyzing social frequency and structure changes in the abnormal time period, and extracting social variation features; a cognitive toughness index is calculated by integrating the health archive and the recovery ability to the health event; and performing toughness weighting on the multi-dimensional features to construct a time sequence tensor, and inputting the time sequence tensor into a recursive model to predict a cognitive impairment risk value. And if the risk value suddenly changes, the system automatically backtracks the feature trajectory of nearly 7 days, constructs and screens a prediction path with the strongest interpretation force, outputs a dominant prediction result and a key factor sequence, and realizes high-interpretability and high-reliability early recognition and intervention reference.

The invention discloses a cascade water-wind-light long-term complementary scheduling rule optimization method under multiple uncertainties. The method comprises the following steps: S1, quantifying water-wind-light multiple uncertainties; s2, constructing a cascade water-wind-light long-term complementary scheduling model; and S3, SDDP-based cascade water, wind and light optimization scheduling solution is carried out. According to the cascade water-wind-light long-term complementary scheduling rule optimization method under multiple uncertainty, a multiple uncertainty quantification method based on a Markov chain and an ARMA model is constructed to represent a long-term runoff and wind-light output random process, and meanwhile, a cascade water-wind-light long-term complementary scheduling model is provided. A nonlinear stage model is converted into a linear stage recursive model in combination with disjunction planning, finally, a cascade water-wind-light optimal scheduling algorithm based on SDDP is put forward for efficient solving, a water-wind-light long-term complementary scheduling rule is deduced, and the problem of cascade water-wind-light long-term regulation and control is solved.

The invention relates to multi-dimensional path geometry tracking for autonomous systems and applications. In various examples, a recursive model may be used to efficiently track and / or predict geometries associated with one or more paths in an environment. For example, the disclosed systems and methods may use Kalman filters to track and predict control points corresponding to Bezier curves (e.g., 2D and / or 3D Bezier curves). The Bezier curve may represent a geometry associated with one or more lanes of the travel surface. In some cases, multiple Bezier curves may be used to represent the geometry of the lane, and multiple Kalman filters may be used to track and predict control points for each Bezier curve. For example, edges of a lane may be represented using a first Bezier curve, and control points of the first Bezier curve may be tracked and predicted using multiple Kalman filters (e.g., one Kalman filter is used for each x, y, or z dimension for a 3D Bezier curve).

The application provides a persistent learning type index system for DRAM-NVM hybrid memory, comprising: a model part: dividing data into sub-data sets, expressing each sub-data set by using a linear regression model, collecting maximum keys covered by each linear regression model, constructing a two-layer recursive model index architecture RMI, indexing local maximum keys, and retrieving a model address to which a target key belongs; a data part: comprising two types of data structures, namely a Node node class and a Buffer buffer block class; a structure type adjustment part: according to the data structure type, being divided into buffer block expansion, buffer block conversion, node conversion and node splitting, and being used for ensuring that the index always remains correct and efficient during the whole working load process. The application solves the problem that the index height is not limited, and ensures that the index can correctly and efficiently access data under any type of working load.

The invention relates to the technical field of space, in particular to an interference measurement constellation error sensitivity analysis method based on a proxy model. The method comprises the following steps: setting a space interference measurement constellation orbit coordinate system and related parameters; constructing a spacecraft orbit recursion model in the interference measurement constellation; constructing a spatial interference measurement constellation configuration stability index model, a 0-order cutting point of the agent model, a first-order item of the agent model, a second-order item of the agent model and an interference measurement constellation error sensitivity agent model; and giving an initial deviation based on an interference measurement constellation error sensitivity agent model, predicting a configuration stability index evolution condition, and analyzing the interference measurement constellation error sensitivity based on the predicted configuration stability index evolution. By adopting the interference measurement constellation error sensitivity analysis method based on the proxy model, the calculation precision and the calculation efficiency of error sensitivity analysis are effectively improved, the calculation mode is simple, the robustness is good, and the order is convenient to expand.

The application belongs to the field of robot model identification, and discloses a compound robot vehicle-arm coordinated dynamics optimal excitation and high-precision identification method. The application designs a non-singular linear regression equation through a dynamics recursive model of compound robot vehicle-arm coordinated movement, then designs an optimal excitation trajectory of vehicle-arm coordinated movement based on a limited term Fourier series and a dynamics regression matrix weighted condition number, and designs a high-precision adaptive weighted iterative solution algorithm of dynamics parameters based on abnormal data suppression and measurement noise prior knowledge, and finally realizes high-precision robust identification of the compound robot vehicle-arm coordinated dynamics model. The method can realize vehicle-arm coordinated full-state dynamics linearization modeling and optimal synchronous excitation, and realize high-precision robust identification of the dynamics model parameters in the face of measurement noise and abnormal data, and finally effectively improves the accuracy and physical consistency of the compound robot model identification.

The present invention discloses a method for designing a conceptual diagram of the appearance and interior decoration of a ship based on the relationship between text and graph, comprising the following steps: S1, performing entity recognition and relationship extraction on the images of the appearance and interior decoration of the ship; S2, based on the neural network architecture of text-graph modeling, splicing and fusing the image feature representation and the text feature representation corresponding to the labels to generate a fused feature representation; S3, fusing the latent space representation self-learned by the variational autoencoder and the image generation ability of the generative adversarial network, optimizing the parameters of the two through joint training, and generating a conceptual design diagram of the appearance and interior decoration of the ship based on the transformation or differentiability of the rules of the recursive model.

The invention relates to a building heat conduction parameter identification method and device, computer equipment and a storage medium, and relates to the technical field of virtual power plants. The method comprises the following steps: acquiring an analytic recursive model of a second-order equivalent heat conduction parameter model of a target temperature control device; the analytic recursive model comprises a heat conduction parameter to be identified; respectively obtaining a size parameter and a state measurement data set of the building space and a current regulation and control parameter of the virtual power plant; determining a target value range of the heat conduction parameter according to the size parameter and the current regulation and control parameter; and determining a target value of the heat conduction parameter according to the state measurement data set, the analytic recursion model, the target value range and a preset termination condition. By adopting the method, the identification accuracy of the building heat conduction parameters can be improved.

The application provides a multi-target ambiguity resolution and redundancy reduction method based on a multi-ary tree recursive model, comprising the following steps: step 1, according to a device repetition period parameter, a residual difference table is established for each two relatively prime repetition periods; step 2, after an apparent distance difference is obtained according to an apparent distance and a table is looked up to obtain an ambiguity, a target real distance is determined; step 3, according to the residual difference table, a matrix type staggered index two-dimensional table is established for a target index, a distance true value and a related repetition frequency index; and step 4, according to the established matrix type staggered index two-dimensional table, a method for obtaining multi-visibility of a target based on a multi-ary tree recursive model is adopted to perform a high-to-low fast search for multi-visibility of a target for all distance unit targets of each repetition period until the target is determined. The application effectively reduces redundancy of multi-target ambiguity resolution and improves processing speed.

This invention discloses an efficient modeling method for electromagnetic transient simulation of modular multilevel converter battery energy storage systems, belonging to the field of power system simulation technology. The method includes: determining multiple operating modes of a submodule in a modular multilevel converter battery energy storage system, and for each operating mode, writing the state equation of the submodule within one cycle; performing state averaging modeling on multiple state equations using a periodic averaging operator to obtain a state-space averaged circuit simulation model of the submodule; and processing the state-space averaged circuit simulation model using a discrete difference method to obtain a state-space averaged discrete analytical recursive model of the submodule. This technical solution can reflect the high-frequency dynamic characteristics affected by the switching process, simulate the unbalanced operating state of the submodule, and improve the simulation speed.

The invention discloses an online fractional order parameter identification method based on a neural network and secondary ripple suppression, and the method comprises the steps: taking a single-phase two-stage inverter system as an object, carrying out the modal analysis of a fractional order model of a front-stage buck circuit, and obtaining a fractional order state equation representing the dynamic characteristics of a capacitor / inductor; establishing a fractional order discretization recursion model of the single-phase two-stage inverter based on GL fractional order calculus definition; constructing a neural network structure according to a recursive relationship between capacitor voltage and current in the discrete model, and constructing a mean square error loss function by taking actual sampling voltage and current values as reference values and taking neural network output as an estimated value in the process of normal operation of the inverter and synchronous suppression of second harmonics; solving a partial derivative of the loss function to obtain a gradient expression corresponding to each weight, and updating the weight of the neural network by adopting a gradient descent algorithm; and calculating and outputting parameters of the fractional order capacitor and the inductor in real time according to the fractional order discrete model and the weight.

This invention discloses a dynamic prediction method for tunnel displacement fields based on a physical information neural network, belonging to the field of tunnel engineering technology, and serving real-time deformation prediction during tunnel excavation. The invention first uses historical excavation data to generate timestamped displacement and strain field sequences, and trains a PINN physical inference model and a PLSTM temporal recursive model respectively. As the actual excavation progresses, the overall displacement and strain are rapidly calculated based on the geological parameters of the current step, and then input into the temporal recursive model to predict the displacement field of the next excavation step. Simultaneously, the mean square value and second derivative of each parameter along the excavation step sequence are analyzed to automatically identify floating and unstable parameters, and four sets of composite geological scenarios are constructed accordingly. These scenarios drive the prediction of the corresponding displacement field set output by the pipeline in parallel, forming a displacement envelope. This invention achieves physically reasonable and risk-quantified dynamic prediction of displacement fields during the excavation process, providing a quantitative basis for early warning and support decisions.

The application provides an interval event matching method and system of a content-based publish / subscribe system, through constructing a predicate disjoint set and an optimized partition model thereof, and integrating a learning index of a range query recursive model index in each partition, high-speed, stable and verifiable matching of events to a subscription set under multi-attribute interval predicates is realized; in particular, the traditional memory-intensive matching is reconstructed into a calculation-intensive process mainly based on model reasoning, while correctness and completeness are ensured, average latency is reduced and long tail delay is inhibited, and key modules such as partition and coverage optimization, model training and online reasoning, error boundary control and limited correction, and accurate back matching of residual set subscription are included.

The present application aims to provide a kind of center difference information filtering phase unwrapping method based on deep learning, comprising the following steps: A, CDIF phase unwrapping recursive model and deep neural network are constructed, the deep neural network is based on LANET interferogram semantic segmentation network;B, after interferogram is input into deep neural network processing, the semantic segmentation image based on interferogram fringe distribution is obtained;C, the merging and optimization adjustment of interferogram semantic segmentation area are carried out, and the segmentation area with appropriate size is obtained, to constitute interferogram area segmentation graph D, each segmentation area in interferogram area segmentation graph is unwrapped using CDIF phase unwrapping recursive model;E, the unwrapping result of each segmentation area is merged, and the unwrapping phase of whole interferogram is obtained.The present application can obtain more robust results in interferogram phase unwrapping experiment, and its efficiency can be accepted, can effectively process interferogram phase unwrapping problem.

The invention relates to an anomaly detection method and system for Internet of Things data and electronic equipment, the method is applied to a processor with a parallel processing function, and the method comprises the following steps: acquiring sensor monitoring data at different moments, acquiring an input sequence according to the monitoring data, the transformation matrix is obtained according to a recursion model of an exponentially weighted moving average method and a preset smoothing factor; performing accumulation according to the input sequence and a preset initial matrix to obtain an output sequence, wherein elements in the output sequence represent an operation result of a moment corresponding to the monitoring data; and in response to the condition that any element in the output sequence does not meet the preset threshold range, judging that the monitoring data is abnormal data.

In various examples, geometries associated with one or more paths in an environment may be efficiently tracked and / or predicted using recursive models. For instance, the disclosed systems and methods may use Kalman filters to track and predict control points corresponding to Bezier curves (e.g., 2D and / or 3D Bezier curves). The Bezier curves may be representative of geometries associated with one or more lanes of a driving surface. In some instances, multiple Bezier curves may be used to represent a geometry of a lane, and multiple Kalman filters may be used to track and predict control points for each Bezier curve. For instance, an edge of the lane may be represented using a first Bezier curve, and control points for the first Bezier curve may be tracked and predicted using multiple Kalman filters (e.g., for a 3D Bezier curve, one Kalman filter for each x, y, or z dimension).

The present application relates to steel pipe bolt loose state prediction technology field, especially in kind based on ultrasonic identification steel pipe bolt loose state prediction method and system. Content includes: obtaining multichannel ultrasonic reflection signal, and projecting to three-dimensional coordinate space, constructing ultrasonic energy attenuation distribution function;Based on ultrasonic energy attenuation distribution function, all multichannel ultrasonic reflection signal is mapped to four-dimensional space-time disturbance tensor field, and decoupling processing is carried out, and the disturbance energy offset rate is obtained;The disturbance energy offset rate is uniformly expressed as state vector, and a nonlinear time series recursive model is constructed to obtain the predicted state vector;Based on the predicted state vector, a double branch evaluation mechanism is introduced, state classification and pretightening force regression prediction are carried out. The traditional steel pipe bolt loose state prediction technology is not accurate for ultrasonic signal processing analysis, the high-dimensional space disturbance characteristics caused by bolt loosening cannot be obtained, which leads to low recognition accuracy and poor stability.

The application relates to the technical field of battery management systems, in particular to a battery capacity online estimation method based on an extended Kalman filter, which mainly comprises the following steps: determining system state variables and system observation variables; establishing a system state equation and a system observation equation; establishing a recursive model for battery capacity estimation according to the extended Kalman filter algorithm; and estimating the battery capacity according to the recursive model after initial information is determined. The application has low error sensitivity to charging and discharging capacity and SOC, can realize online estimation of the battery capacity in the actual use process of the battery, and can break the dependence on the initial capacity of the battery after a certain number of iterations.

The invention provides an interval event matching method and system of a content-based publishing / subscribing system, and the method comprises the steps: constructing a predicate disjoint set and an optimized partition model thereof, and integrating a learning index of a range query recursive model index in each partition; high-speed, stable and verifiable matching from the event to the subscription set under the multi-attribute interval predicate is realized; in particular, traditional memory intensive matching is reconstructed into a calculation intensive process taking model reasoning as a main part, correctness and completeness are guaranteed, meanwhile, average time delay is reduced, long-tail delay is restrained, partitioning and coverage rate optimization, model training and online reasoning and error boundary control and finite correction are included, and the calculation intensive process is optimized. And key modules such as precise backspacing matching of residual set subscription.

In various examples, geometries associated with one or more paths in an environment can be efficiently tracked and / or predicted using recursive models. For example, the disclosed systems and methods can use Kalman filters to track and predict control points corresponding to Bézier curves (e.g., 2D and / or 3D Bézier curves). The Bézier curves can represent geometries associated with one or more lanes of a road surface. In some cases, multiple Bézier curves can be used to represent a lane geometry, and multiple Kalman filters can be used to track and predict control points for each Bézier curve. For example, a lane edge can be represented using a first Bézier curve, and control points for the first Bézier curve can be tracked and predicted using multiple Kalman filters (e.g.,(for a 3D Bezier curve, a Kalman filter for each x, y or z dimension).

The invention discloses a method for judging the bonding strength of a new and old concrete interface based on Deepseek, and relates to the technical field of artificial intelligence type discriminating.The method comprises the steps that information acquisition is adopted, specifically, variable information of the bonding strength of the new and old concrete interface is acquired, and the variable information further comprises influence factor information influencing the bonding performance; factors of bonding strength of new and old concrete interfaces are extracted through an artificial intelligence Deepseek detection algorithm and are analyzed, and then levels of an overall architecture, functions of each layer, training and reasoning processes from input to output, and cooperative work of different components are summarized by using a recursive model of deep learning and a neural network. And data flow, hierarchical feature extraction and optimization target development are performed, and diversified tasks are adapted through different components, so that the bonding strength of the new and old concrete interface is analyzed and judged in a comprehensive understanding and preferential manner. The judgment method has remarkable advantages, influence characteristics and influence factor information variables can be automatically and efficiently extracted, and tasks can be flexibly processed, so that the workload of a user is reduced, and the work of judging the bonding strength of the new and old concrete interface is improved.

A method for online remaining life prediction of machinery using a recursive model update strategy first establishes a degradation model and model library for the mechanical equipment, then determines three thresholds and, based on the thresholds, determines the starting degradation time and the starting prediction time. Model parameters are then updated using a maximum likelihood estimation-expectation maximization combined algorithm based on a sequence of observations after the starting degradation time, and the optimal degradation model is automatically selected using the recursive Bayesian Information Criterion (RBIC). Finally, the remaining life prediction result is obtained based on the selected model and its updated parameters. The method is driven by the mechanical equipment's own condition monitoring data and uses the RBIC score to match the monitoring data to the optimal degradation model in real time. This method effectively characterizes the degradation of the mechanical equipment during operation in industrial practice and improves the accuracy of the remaining life prediction of the mechanical equipment.

This invention relates to the field of remote sensing technology and discloses a remote sensing estimation method for ecological carrying capacity. The method includes: constructing a standard spatiotemporal benchmark for multi-source data fusion; using surface temperature as a nonlinear correction factor to adjust the gain of nighttime light and calculating a human activity pressure index; calculating the neighborhood thermal potential difference based on the thermodynamic gradient principle and converting it into a thermal radiation inhibition coefficient; combining water deficit and thermal radiation inhibition to correct and normalize potential productivity, obtaining a standardized effective ecological supply index; constructing a time-series recursive model with a missing compensation mechanism to calculate ecological fatigue; and finally, comprehensively calculating the ecological carrying capacity state by integrating the supply index, pressure index, and ecological fatigue. This invention solves the problems of nighttime light saturation overflow, inconsistent supply and demand dimensions, and lack of consideration for historical cumulative effects, achieving a dynamic and refined assessment of regional ecological carrying capacity.

The present invention provides a passive collaborative tracking method based on Sage-Husa adaptive filtering, which constructs a system recursive model and an observation model, obtains the collaborative positioning target position as the initial value of the system state; generates a Sigma point set, obtains a one-step prediction value of the state quantity and the state covariance; generates a state prediction Sigma point set, obtains the measurement prediction value and measurement covariance of the sensor; obtains the residual amount, and stores the residual amount in a storage pool; determines whether the current storage pool is full, and if the current storage pool is full, estimates the measurement noise according to the residual amount to obtain the measurement noise covariance, and updates the measurement covariance; if the current storage pool is not full, updates the measurement covariance according to the initialized measurement noise covariance; obtains the filter gain, updates the state quantity and state covariance, enters the next moment, and returns to generate the Sigma point set according to the state quantity at the previous moment. The technical solution of the present invention can solve the technical problem of limited passive collaborative tracking performance in the prior art.

The application discloses a dynamic self-adaptive energy consumption optimization method and device for casting, aiming at the problem that the existing energy consumption management is static and difficult to be optimized online under the constraints of process, safety, beat and maximum demand due to the multi-process coupling and working condition fluctuation of casting, the device set and metering, state acquisition and instruction issuing interface are constructed, the time sequence data is synchronously acquired and the abnormality is marked, the process coupling relation graph of three types of relations is established and the edge weight is dynamically updated, the device state recursive model is constructed, the error boundary is formed by conformal prediction to form the risk margin constraint, the conditional diffusion control instruction generation model is trained, the output is constrained by the behavior support domain, the executable control instruction sequence is generated by combining the control barrier function filtering and multi-step constraint checking, and the execution and rewriting update are issued. The application is used for the full-process collaborative energy consumption optimization control of casting production, reduces the comprehensive energy consumption and peak energy consumption, and improves the operation stability.

This invention relates to the field of parameter field cascade reconstruction technology, and discloses a method for cascade reconstruction of three-dimensional water vapor parameter fields based on sparse GNSS station network data. The method includes: 1. Collecting observation data from m GNSS stations sparsely distributed within a target area; 2. Obtaining the slant path water vapor content of all GNSS signals corresponding to each GNSS station; 3. Establishing a water vapor tomography equation set for each GNSS station, using a single GNSS station as the basic unit; 4. Obtaining the local three-dimensional water vapor parameter field of each GNSS station; 5. Dividing the entire target area into spatial units with the same spatial resolution as the local three-dimensional water vapor parameter field of each station, and calculating the spatial correlation coefficient between each spatial unit and the local three-dimensional water vapor parameter field of each station at each height level; 6. Constructing a recursive model to obtain the three-dimensional water vapor parameter field of the target area. This invention achieves high-precision reconstruction of the three-dimensional water vapor parameter field of the target area under sparse GNSS station network conditions, improving the inversion accuracy of three-dimensional water vapor parameter fields based on sparse GNSS station networks.

Provided is at least one processor included in an information processing apparatus, carrying out: a feature map generation process of generating a plurality of scale-specific feature maps from input data; a feature series generation process of generating a feature series from the plurality of scale-specific feature maps; and a feature information generation process of generating feature information by inputting the feature series into a recursive model.

The present invention provides a data-driven control method for a precision positioning platform based on skipped inversion, comprising the following steps: S1, establishing an autoregressive model with time-varying linear parameters and external inputs based on the dynamic characteristics of the precision positioning platform; S2, designing an adaptive parameter estimator based on a dynamic regression extension and hybrid method; S3, constructing a p-step recursive model of the precision positioning platform based on the adaptive parameter estimator; and S4, designing an adaptive controller based on skipped inversion based on the number of zero points outside the unit circle of the autoregressive model and the estimated parameter values. This method primarily addresses the problems of real-time parameter estimation and unstable inversion faced by precision positioning platform tracking control. It can estimate the true parameter values ​​in real time when the parameters do not change drastically, overcome the control divergence problem caused by unstable inversion, and improve the tracking performance of the precision positioning platform.

This invention relates to the field of computer vision technology, specifically providing a video 3D human pose estimation method based on a state-space recursive model. The method first extracts human keypoint information from video frames using a 2D pose estimator, and encodes it into frame-level features via a pose embedding module. Then, a state-space recursive network is used to achieve temporal modeling with linear time complexity. A token pruning clustering module based on state density peaks is combined to select keyframes. Representative tokens are input into the state-space recursive network for long-sequence temporal modeling with linear computational complexity. Finally, state interpolation and propagation mechanisms are used to recover the complete sequence, achieving 3D human pose reconstruction. This method has low computational complexity, strong real-time performance, and is suitable for video pose estimation tasks in edge computing and embedded devices.