17 results about "Minimization algorithm" patented technology

The invention relates to the technical field of artificial intelligence and energy system operation control, and discloses an artificial intelligence-based energy dynamic aid decision-making method and system, and the method comprises the steps: collecting multi-source heterogeneous data of an energy system, and constructing an observation vector; decoupling the observation vector into an environmental causal factor and an environmental noise factor of a potential space by using a variational structured causal characterization model; pruning a causal factor based on an accumulated anti-fact regret value gradient to obtain effective decision information, generating an anti-fact trajectory by applying intervention to a noise factor to construct an expanded belief state set, and updating a control strategy by using a continuous anti-fact regret minimization algorithm on this basis; and finally, performing physical constraint verification on the sampling action and executing the physical constraint verification. Through causal decoupling and an anti-factual reasoning mechanism, decision robustness, interpretability and physical safety of an energy system in a dynamic non-stationary environment are effectively improved.

A modelling method for making a virtual model of a user's head includes operations of: acquiring a plurality of depth images of a user's head in different head poses; processing the acquired depth images; localising on the two-dimensional images first marker points corresponding to characteristics of the head; estimating the head by estimating the yaw, pitch and roll angles of the head and the positioning of the head along three axes of a reference system of Cartesian axes for each of the two-dimensional images of the depth images, performing a transformation of the head pose; projecting the first marker points located in the respective cloud of points; associating iteratively in pairs the clouds of points until obtaining a cloud of synthesis points; aligning the cloud of synthesis points; and deforming the statistical head model by means of a minimization algorithm of a predefined cost function.

The application discloses an evolution modeling and prediction method and system for extreme risk coupling of an aviation hub network, and the method comprises the following steps: acquiring static section data of an aviation transportation system; constructing a space coupling dynamics model according to the static section data, wherein each airport node in the space coupling dynamics model corresponds to an attraction weight, and the attraction weight represents the co-mobility characteristics of each airport node under an extreme load state; using a maximization minimization algorithm to inverse the attraction weight; based on the inverted attraction weight, acquiring a space core parameter group through a Markov chain Monte Carlo algorithm, and obtaining regional congestion co-mobility early warning information based on the space core parameter group. The application can accurately simulate the dynamic growth logic of a multi-center airport group, and can more quantitatively identify the cascade load pressure distribution within the space dependence under the extreme scenario of the failure of a specific core hub.

This invention discloses a method for 3D modeling of small celestial bodies based on SFS (Self-Focused Spectroscopic Method), belonging to the field of intelligent 3D modeling and image reconstruction technology for small celestial bodies. To improve the accuracy of 3D modeling of small celestial bodies, this invention performs modeling under the condition that the camera's extrinsic parameters are determined. This includes setting the positions of the camera and the small celestial body in a simulation environment and setting the camera's motion trajectory to obtain multiple frames of small celestial body observation images; extracting the contours from the small celestial body observation images; transforming the 3D modeling problem of small celestial bodies into a pseudo-Boolean function minimization problem; applying local minimization and hierarchical minimization algorithms to solve for the minimum value of the pseudo-Boolean function to obtain a 3D voxel model of the small celestial body. For modeling under the condition that the camera's extrinsic parameters are uncertain, this includes converting the 3D voxel model of the small celestial body into a polygon model; selecting 3D model points; matching the 3D model points with 2D image points; and applying the LM algorithm to optimize the matching error between the 3D model points and 2D image points. This patented method is applied to the process of small celestial body detection.

The invention relates to the technical field of radar observation, and discloses a radar observation intelligent collaborative assimilation method, which comprises the steps of constructing a Python-Fortran hybrid programming platform, commanding a weather forecast mode simulation environment, generating multi-scale background field data, and enabling the programming platform and a weather forecast mode to share data mutually. A variational auto-encoder and a decoder are used to realize rapid mutual conversion between a physical space and a low-dimensional hidden space, the encoder realizes data dimension compression through a neural network, and the decoder realizes data physical reconstruction through the neural network; in the hidden space, radar observation is fused into a mode background by using a minimization algorithm to realize multivariable nonlinear assimilation. The physical environment state after assimilation is output to form a continuous time sequence which can be used for three-dimensional environment monitoring service, and an assimilation result at the latest moment is used as an initial state of a weather forecasting mode, so that the weather subjected to numerical simulation naturally expands to the future in a time domain, and forecasting service can be formed. According to the method, the efficiency and precision of three-dimensional environment monitoring and forecasting are improved.

The invention provides a self-anti-S box construction method and system based on m-sequence generation and logic optimization, and belongs to the technical field of cryptography and hardware security. The method comprises the following steps: firstly, constructing an n-level linear feedback shift register according to a set S box number n and a primitive polynomial, generating an m sequence in a non-all-zero initial state, and adding an all-zero state at the tail; a self-reverse S box lookup table is constructed through a symmetric mapping strategy, so that the ith element in the sequence and the element at the symmetric position form mapping, and the same S box is used for encryption and decryption. Thirdly, the lookup table serves as a Boolean function truth table, is simplified into a simplest expression through a logic minimization algorithm and is decomposed into a layered logic circuit; during actual processing, N data blocks are converted into a bit plane format from block type storage, then parallel operation is carried out through a hierarchical circuit, and finally a result is transposed back to block type storage. According to the method, the same S box hardware module can be reused in the encryption and decryption process, and hardware resources are remarkably saved.

The invention belongs to the technical field of circuit fault diagnosis, and particularly relates to a solving method and device based on multiple observation aggregation fault diagnosis and a medium. In order to solve the problems of one-sided single observation information, redundant diagnosis results and the like, multiple times of fault observation are performed on a circuit system, and multiple independent diagnosis results are generated and integrated into a diagnosis set; establishing a mapping relation between a diagnosis set and a unique keyword by using a map function of the STL, and converting the component combination into an integer identifier to simplify operation; calling a BWSS algorithm to perform collision calculation on the mapped keywords to obtain a candidate solution set covering all observation scenes; after the keyword solution set is converted to an original diagnosis form, deleting a superset according to a redundancy rule to form a simplified intermediate diagnosis solution set; and finally, through a branch and bound minimization algorithm, according to potential classification and ascending order arrangement of the diagnosis sets, high-order supersets are deleted layer by layer by adopting a heuristic strategy, and finally, a minimum diagnosis solution set of all faults is output.

The application discloses an industrial surface defect detection method and system based on sampling disturbance optimization and history reuse, and relates to the technical field of industrial visual defect detection. In view of the problems that the background of an industrial defect image is complex, the loss topography is highly non-convex, which leads to model overfitting, and the traditional sharpness perception minimization algorithm relies on double back propagation, which leads to excessive algorithm cost, a lightweight optimization method is provided. The method introduces a sampling sharpness estimation mechanism in the network parameter update stage, generates a disturbance tensor by sampling within the preset norm boundary to explore the flat minimum value, and introduces a history reuse mechanism to dynamically evaluate and reuse the historical state, so as to bypass the back propagation calculation of the current calculation graph in part of the training period. With less memory cost, the application greatly reduces the disturbance calculation cost, effectively crosses the non-convex loss valley, significantly improves the generalization accuracy of industrial surface defect detection, and has a wide application prospect.

This invention discloses a semi-blind direct localization method for thermal noise communication systems, comprising: based on the frequency domain received signal model of the thermal noise communication system, utilizing the statistical characteristics of the received signal to construct a maximum likelihood estimation problem concerning the mobile station position and channel fading coefficient, obtaining an objective function containing logarithmic determinant terms and trace function terms of a matrix; the objective function simultaneously includes a contribution from pilot power and an expected contribution from the power of the received data signal; using an optimization minimization algorithm framework, alternately solving for the mobile station position and channel fading coefficient until a preset convergence condition is met, obtaining the localization result; wherein, by constructing a substitution function containing the statistical characteristics of blind pilots and data, a closed-form solution for the channel parameters is derived in each iteration, and the mobile station position is updated through a two-dimensional grid search. This invention mines the statistical information of the data portion, effectively improving the utilization rate of the received signal and increasing the localization accuracy under low signal-to-noise ratio conditions.

The application discloses an improved traveling wave distance measurement method for fault location analysis of a power distribution network, and relates to the technical field of power distribution networks. The method solves the problem of positioning deviation caused by a traditional single wave speed model, and especially the problem of false fault points caused by refraction and reflection at the connection between a cable and an overhead line. The method performs waveform translation and coincidence verification, calculates the amplitude difference average under different translation amounts, automatically identifies the optimal time alignment point, and locks the associated time difference t1. This process can effectively correct the time error caused by signal distortion, non-synchronous sampling and the like, and can improve the time difference accuracy from the microsecond level to the nanosecond level, thereby laying a foundation for subsequent accurate positioning. In the locked range interval, a plurality of associated points are randomly selected, a corrected waveform is constructed based on the attenuation ratio, and repeated verification is performed. The optimal solution is selected through a characteristic mean minimization algorithm. The mechanism can effectively suppress single measurement point error, and can improve the robustness of the positioning result through statistical average effect.

The application provides a method for realizing safe communication by STAR-RIS assisted movable antenna, and relates to the technical field of wireless communication. The method maximizes the safety and rate of the system by jointly optimizing the base station beam forming, the movable antenna position and the STAR-RIS coefficient matrix. Specifically, the method comprises the following steps: constructing a downlink MIMO system and establishing a joint optimization problem; using an alternating optimization framework to decompose the joint optimization problem into three sub-problems; using a particle swarm optimization algorithm, a weighted least mean square error method combined with a Majorization-Minimization algorithm to solve each sub-problem; and alternately performing relevant steps until the system safety and rate converge, to obtain the solution of the joint optimization problem. The application uses the above method, significantly improves the system safety communication performance, realizes full space coverage and accurate beam control, and has strong algorithm convergence and high solution efficiency, and is suitable for a downlink system with multiple users and multiple eavesdropping nodes.

The invention belongs to the field of image processing circuits, and particularly provides a hyperspectral image target detection circuit based on a constraint energy minimization algorithm innovatively from the perspective of an analog circuit, so that one-step hyperspectral image pixel point target detection of a hardware circuit is realized, and the defects in the prior art are overcome. The built integral circuit model can be divided into two parts, one part is an optimal filter coefficient calculation module realized through an integral negative feedback circuit, and the other part is an FIR linear filter module realized through a multiply-accumulate circuit. According to the designed circuit model, the actual calculation effect of the pixels of the hyperspectral image data set can be further studied if the spectral information of the target is known in the target detection process. The circuit can be applied to scenes such as remote sensing monitoring and environment detection, hardware implementation of the circuit has important engineering application value and development potential, and core intellectual property right can be formed in the field.

A patch-based deblurring method to deblur images blurred in a non-uniform manner (e.g., motion blur or camera shake) uses a modified total-variation (TV) minimization algorithm to make a network specifically designed to handle non-uniform blur.

The invention discloses a multi-component collaborative vibration monitoring and fault positioning system for a transmission chain of a wind turbine generator, and relates to the technical field of state monitoring and fault diagnosis of wind power generation equipment. An improved order analysis method based on second-order fitting and absolute value nearest neighbor matching is adopted to obtain a stable order spectrum resisting rotation speed fluctuation; after an observation feature matrix is constructed, an SCAD penalty function is introduced to establish a sparse non-negative matrix factorization model, efficient and stable solution is carried out by adopting a near-end alternating linearization minimization algorithm, and accurate decoupling of fault features of multiple components such as a main bearing, a gearbox and a generator is realized; and finally, accurate fault positioning and identification are realized through cosine similarity matching. According to the method, the problems of coupling, aliasing and feature extraction of fault signals of multiple parts under the variable rotating speed are effectively solved, and the diagnosis accuracy, the coverage range and the positioning precision are remarkably improved.

The invention relates to the technical field of hydraulic machinery operation prediction and control, in particular to an Informer-based pump turbine pressure pulsation prediction method, which comprises the following steps of: obtaining a real-time pressure pulsation signal acquired by a pressure sensor arranged on a pump turbine, and sequentially carrying out median filtering method denoising and normalization preprocessing on the real-time pressure pulsation signal; a whale optimization algorithm WOA is adopted to optimize VMD feature parameters based on a signal reconstruction error minimization algorithm; a whale optimization algorithm WOA is adopted to optimize Informer model parameters based on a signal reconstruction error minimization algorithm, target parameter combination is adopted to predict each IMF component, and all predicted IMF components are reconstructed; and calculating a prediction result index, and performing an Informer model adaptive updating mechanism or re-optimization. According to the method, the problems that the performance of a prediction model is unstable and the prediction model is caught in local optimum in the prior art are solved.

This invention discloses a hybrid beamforming method for integrated MIMO communication and radar based on overlapping subarrays. Under the conditions of hybrid analog-digital arrays, including overlapping subarray structure, transmit power, and constant mode constraints, a joint problem of maximizing communication spectral efficiency and minimizing radar beam pattern matching error is established. This joint problem is relaxed into a weighted summation minimization problem, where the designed hybrid beamformer minimizes the Euclidean distance to the optimal communication / radar beamformer. This invention decomposes the hybrid beamforming design problem into three subproblems and develops an efficient alternating minimization algorithm. The proposed hybrid beamforming design based on overlapping subarray structure offers flexibility and effectiveness in terms of spectral efficiency and radar beam pattern performance.

The application discloses a game interaction method and device based on a policy network, equipment and a medium, and relates to the technical field of computers, which comprises the following steps: generating a to-be-trained sample through a sub-process, a local network copy and a deep counterfactual regret minimization algorithm in a preset parallel training architecture; the local network copy comprises a copy of a policy network and a copy of an advantage network of each player; writing the to-be-trained sample into an experience buffer in batches through a main process in the architecture; training the policy network and the advantage network by using a preset multi-dimensional hierarchical sample sampling strategy and the sample in the experience buffer through the main process, obtaining to-be-updated network parameters; updating the network according to the to-be-updated network parameters through the sub-process, generating the to-be-trained sample again based on the updated local network copy, and stopping the training until the training is terminated; and generating a current action decision for the player based on the trained target policy network, so that the player can interact with other players. The application effectively realizes parallel training for DeepCFR.