20 results about "Distributed filtering" patented technology

A length of optical fiber suppresses a nonlinear optical process so as to inhibit energy transfer away from a desired laser wavelength. The fiber has a core and a cladding designed to propagate a laser beam at the desired wavelength, alongside a nonlinear laser component having intensity that escalates as a function of distance due to the nonlinear optical process; a series of spaced-apart filters, each configured with a transmission level to redirect a proportion of the nonlinear laser component from the core into the cladding so as to collectively control exponential growth in the intensity of the nonlinear laser component by imparting resets in the intensity at predetermined intervals along the length; and an output configured to provide the laser beam following its suppressed escalation of the intensity and preservation of energy of the laser beam at the desired wavelength.

The invention belongs to the field of power supply distribution optimization, and particularly relates to an adversarial learning-based power supply distribution unit multi-terminal cooperative correction system, which comprises an acquisition module for realizing differential data acquisition and anomaly recognition through an enhanced acquisition adjustment model, and a preprocessing unit for performing noise reduction and sparse processing by adopting a distributed filtering model under a federated framework; the self-calibration module predicts temperature change through a distributed LS-SVR generator, predicts voltage and current drift amount by combining a BP neural network discriminator of particle swarm optimization, performs adversarial training by using a comprehensive adversarial loss function, and dynamically adjusts a real-time demand distribution coefficient to enable the deviation between the real-time demand distribution coefficient and a true value to meet a threshold value; according to the invention, the collection efficiency, the correction precision and the adaptability to the dynamic change of the complex load are remarkably improved.

The application relates to the technical field of all-in-one machine control, in particular to a five-axis drive control all-in-one machine control method, system, equipment and medium. The application detects a power supply state through a power management module and performs filtering, each separated inverter module is initialized in parallel to obtain an enable signal; target position instructions are generated based on user input and dynamic constraint parameters, real-time position and speed instructions are calculated; a five-axis servo system is driven to execute movement and feedback adjustment; current fluctuation characteristics and position deviation change rates are monitored in real time, when a source abnormality is detected, only the inverter module corresponding to the abnormal axis is turned off; after replacing the abnormal module, the parameters are automatically loaded and the inter-axis synchronization compensation is optimized through adaptive configuration. The application reduces electromagnetic interference through distributed filtering, reduces maintenance cost through modular fault isolation, and realizes system reliability through intelligent adaptive control.

The invention relates to the technical field of artificial intelligence and nutrition and health, and discloses a total nutrition module customer marketing distributed filtering method and system.The total nutrition module customer marketing distributed filtering method comprises the steps that a nutrition and health causality graph is constructed, learning a causal relationship between nutrient elements and health indexes from historical nutrition and health data by using a causal discovery algorithm; realizing a dual Markov decision process, and constructing the dual Markov decision process; an anti-fact effect evaluation system is established, and long-term causal effects of different nutrition intervention paths are evaluated; carrying out personalized intervention effect estimation, and carrying out personalized estimation on the nutrition intervention effect of each customer by utilizing a causal machine learning technology; a distributed calculation and decision execution system is constructed, and efficient processing of large-scale customer data and real-time execution of recommendation decisions are achieved; according to the method, simple correlation analysis is replaced by causal inference, the individual heterogeneity effect is considered, and the accuracy of nutrition scheme recommendation is improved.

The embodiment of the invention provides a multi-target cooperative tracking design method and related equipment, and belongs to the technical field of unmanned aerial vehicle cluster sensing and multi-target tracking. The method comprises the following steps: cooperatively acquiring visual observation data and TDOA observation data of a target through a plurality of unmanned aerial vehicle nodes; each node processes the two types of data by using a visual PHD filter and a TDOA PHD filter to obtain respective target probability density functions; fusing the two density functions by adopting a dynamic weighting strategy based on confidence to generate unified target probability density distribution; on the basis of the fused density distribution, dividing a Voronoi sub-region for each node and calculating the target density gravity center of the Voronoi sub-region; and each node is guided to move towards the center of gravity by using a Lloyd algorithm, so that cooperative tracking and area coverage are realized. According to the method, the advantages of high precision of vision and wide area coverage of TDOA are fused, the communication burden is reduced by adopting a distributed filtering architecture, and the tracking precision, robustness and efficiency of the system are remarkably improved through adaptive fusion and cooperative control.

This invention belongs to the field of power distribution optimization, and particularly relates to a multi-terminal collaborative correction system for power distribution units based on adversarial learning. The system includes an acquisition module that achieves differentiated data acquisition and anomaly identification by strengthening the acquisition adjustment model; a preprocessing unit that uses a distributed filtering model under a federated framework for noise reduction and sparsification; and a self-calibration module that predicts temperature changes using a distributed LS-SVR generator, predicts voltage and current drift using a particle swarm optimization BP neural network discriminator, and uses a comprehensive adversarial loss function for adversarial training to dynamically adjust the real-time demand allocation coefficient so that its deviation from the true value meets a threshold. This application significantly improves acquisition efficiency, correction accuracy, and adaptability to complex load dynamic changes.

The invention provides an anti-eavesdropping distributed tracking detection method for a target tracking system under low communication resources, and belongs to the technical field of distributed tracking detection. The method comprises the following steps: establishing a random dynamic model of a target tracking system under low communication resources; then designing a time-varying distributed filter under the meaning of minimum mean square error according to the dynamic model; designing a filter gain matrix of each sensor node in the system at the moment; substituting the filter gain matrix into a filter to obtain state estimation of each sensor node in the system at the moment; calculating the upper bound of the estimation error covariance of each sensor node in the system at the moment; and repeating the steps until the requirements are met. According to the method, the problem that the anti-eavesdropping distributed tracking detection of the target tracking system with a periodic scheduling mechanism and an encryption and decryption mechanism is difficult to process at the same time by the existing distributed tracking detection method is solved, so that the tracking performance of the tracking system is improved.

The application provides a distributed graph filtering method, device and computer readable storage medium, wherein the method comprises: in the current iteration process of a network graph, obtaining an input signal of the network graph in the current iteration process and an input signal of each vertex of at least two vertices in the network graph; obtaining an output signal of each vertex of the at least two vertices in the last iteration process; weighting the input signal of each vertex and the output signal of the adjacent vertex in the last iteration process according to a distributed filtering algorithm of an ARMA graph filter to obtain the output signal of each vertex in the current iteration process; determining an output signal of the network graph according to the output signals of the at least two vertices in the current iteration process and the input signal of the network graph; and determining the output signal of the network graph as a filtering signal of the network graph when the output signal of the network graph meets a preset convergence condition.

The invention relates to the field of privacy protection query schemes, in particular to an indoor multi-user personalized space keyword privacy query method based on a GIS enhanced index tree. According to the requirements of multiple users for space keyword privacy query in an indoor environment, encryption indexing is carried out on space and text information of an indoor object by constructing a GISEnhancedIndexTree index structure fusing a Hilbert code, an R tree, a densely distributed filter and an EASPE encryption vector. In the query process, a user request is encrypted into a trap door, and the server sequentially executes space, floor and keyword pre-filtering, encryption vector similarity calculation by means of EASPE, threshold filtering, score aggregation and sorting in the index tree and then returns an encryption result to the user. And meanwhile, a comprehensive position vector construction and local coordinate system-based privacy protection coordinate conversion method is designed, so that the privacy security of the whole flow of data is ensured. According to the method, the efficiency and accuracy of indoor multi-user personalized spatial keyword query are improved while the user privacy is effectively protected.

The utility model discloses a compressor high-efficiency filter device with distributed arrangement and compact structure, which comprises a plurality of groups of filter cartridges arranged on a machine cover, the filter cartridges are communicated with an exhaust cavity in a compressor main body, the plurality of groups of filter cartridges are uniformly distributed on the machine cover, and the filter cartridges are communicated with the exhaust cavity in the compressor main body. The filter cartridge comprises a supporting box penetrating through the machine cover, a filter screen is arranged in the supporting box in a sleeved mode, an installation end connected with the machine cover is arranged at the end of the filter screen and used for efficient distributed filtering of the filtering device, and an automatic cleaning piece penetrating through the filter screen is rotationally arranged on the supporting box. A cleaning end attached to the supporting box and the filter screen is arranged on one side of the automatic cleaning piece, and stress blades for driving the cleaning end are arranged on the other side of the automatic cleaning piece. The efficient filtering device of the compressor is arranged in a distributed structure in the compression process, and has the advantages of being compact in structure, efficient in filtering and automatic in cleaning.

The application discloses a kind of optimization distributed filtering method of target tracking system under work cycle scheduling strategy, the method is as follows: one, the dynamic model of target tracking system with state constraint and probability quantization is established;Two, distributed filter design is carried out to dynamic model under work cycle scheduling;Three, the upper bound of one-step prediction error covariance matrix is calculated;Four, filter gain matrix is calculated;Five, it is obtained in distributed filter by substituting, whether the total length of sensor network is reached, if, then execute six, if, then end operation;Six, the upper bound of filtering error covariance matrix is calculated;Let, execute two, until meet. The application solves the problem that existing distributed filtering method cannot simultaneously process sensor network with state constraint, probability quantization and work cycle scheduling, leading to the problem of reduced filtering performance.

The application discloses a kind of flexible connection constraint under multi-node distributed filtering estimation method, belong to deep space exploration technical field.The implementation method of the application is as follows: by selecting the node with maximum observability in adjacent node, the state estimation thereof is utilized to establish flexible connection constraint, and the constraint least square problem of distributed filtering estimation is formed;The condition that inequality constraint is converted into equality constraint needs to be satisfied is analyzed using KKT condition, the flexible connection constraint is converted from inequality constraint to equality constraint, to avoid optimization process under the influence of inequality constraint to fall into local optimum, by constraint local linearization, the analytical expression of node state estimation mean value and estimation error covariance is obtained, the efficient estimation of flexible lander node state is realized, and the state estimation result conforming to the physical characteristics of flexible connection is generated;In addition, the nonlinear constraint linearization error is reduced by iterative estimation, and the node state estimation accuracy of flexible lander is further improved.

The application provides a target tracking system anti-eavesdropping distributed tracking detection method under low communication resources, and belongs to the technical field of distributed tracking detection. The application establishes a random dynamic model of a target tracking system under low communication resources; then, a time-varying distributed filter under the least mean square error is designed according to the dynamic model; then, a filter gain matrix of each sensor node in the system at a time point is designed; the filter gain matrix is substituted into the filter to obtain the state estimation of each sensor node in the system at the time point; and the estimation error covariance upper bound of each sensor node in the system at the time point is calculated; let, and repeat until the condition is met. The application solves the problem that the existing distributed tracking detection method is difficult to simultaneously process the target tracking system anti-eavesdropping distributed tracking detection problem with a periodic scheduling mechanism and an encryption and decryption mechanism, thereby improving the tracking performance of the tracking system.

The invention provides a sensor network distributed filtering method and system based on adaptive event triggering under replay attack and topology switching, and relates to the technical field of signal processing. The invention aims to solve the problem that the estimation performance is reduced when the existing distributed filtering method has the problems of communication resource limitation, topology random switching, replay attack and the like in a sensor network. The method is characterized by comprising the following steps: establishing a state space model and a sensor output model of a discrete nonlinear time-varying stochastic system, and adopting a self-adaptive event triggering mechanism to enable the measurement of a distributed filter to be available; a replay attack compensation mechanism is established, and the influence caused by the replay attack on the sensor network is processed in time; the developed adaptive event-triggered distributed filter design algorithm is in a recursive form and is very suitable for online application. The method can effectively cope with multiple challenges such as bandwidth limitation, topology change and network security threats, has good applicability and practicability, and is more suitable for practical engineering application.

The invention discloses a distributed filtering method and system based on depth expansion and a storage medium, relates to the technical field of sensor networks, and aims to solve the problems of poor multi-noise adaptation, insufficient node estimation consistency and high calculation complexity of a traditional method. The method comprises the following steps: establishing a network topology and a system model, designing an information fusion filter structure, mapping a system iteration equation into a deep expansion neural network, constructing a composite loss function containing an estimation error and node divergence, and training and solving an optimal gain by using the deep neural network. The method is adaptive to a multi-noise scene, realizes double optimization of precision and consistency, is low in calculation complexity and high in robustness, and is suitable for the fields of environment monitoring, dynamic target tracking and the like.

An unmanned aerial vehicle formation cooperative navigation distributed filtering method applying an event-driven mechanism comprises the following steps: performing local structure division on each unmanned aerial vehicle platform and a neighbor unmanned aerial vehicle platform according to a communication topology in an unmanned aerial vehicle formation cooperative navigation configuration; according to the method, the state cognition difference of neighbor unmanned aerial vehicle platforms to a target unmanned aerial vehicle platform is used as a communication triggering condition, relative measurement and relative state information between the unmanned aerial vehicle platforms are utilized, and after posteriori state estimation of at most an inner neighbor number is obtained after information fusion, filtering prediction updating is carried out. According to the invention, effective release of communication resources of the unmanned aerial vehicle formation applying the distributed collaborative navigation system is realized, communication power consumption is greatly reduced, and robustness, stability and reliability of the navigation system are effectively ensured on the basis of state cognitive consensus among the unmanned aerial vehicles under the condition that communication is reduced.

This invention discloses an adaptive closed-loop control method and system for a continuous stirred tank reactor, relating to the field of continuous stirred tank reactor control technology. The method acquires a preset homogeneous evolution model; performs multi-rate sampling through target nodes to obtain monitoring feature data; uses an attack assessment model to identify random spoofing attack characteristics of the communication channel and performs cleaning and noise reduction to obtain a measured data stream; calculates the data deviation, and when it exceeds a preset evolution threshold, uses the measured data stream as trigger data; inputs the trigger data into a preset distributed filter, solves the local filter gain to generate local state estimates, and weights and fuses these estimates to obtain global state estimates; finally, it compares the estimated state with a preset target state value to generate a state deviation signal, corrects the local filter gain, and generates an execution compensation control command. This invention effectively overcomes interference from multi-rate sampling, network attacks, and bandwidth limitations, achieving high-precision and robust adaptive closed-loop control of the reactor.

An anti-eavesdropping distributed fusion filtering method for the multi-rate nonlinear system over sensor network includes: Step 1, establishing a dynamic model for the multi-rate nonlinear system over sensor network; Step 2, transforming the multi-rate nonlinear system dynamic model into a single-rate nonlinear system dynamic model through the prediction compensation strategy; Step 3, designing an anti-eavesdropping distributed fusion filter; Step 4, calculating an upper bound on the one-step prediction error covariance (tk+1|tk); Step 5, deriving the local distributed filter parameter Ki(tk+1); Step 6, deriving the selection matrix Lij(tk+1); Step 7, substituting Ki(tk+1) and Lij(tk+1) into Step 3 to obtain the fusion filter {circumflex over (x)}CI(tk+1|+tk+1); Step 8, solving for the upper bound on the local filtering error covariance (tk+1|tk+1). The method solves the problem that the existing fusion filtering method cannot simultaneously deal with the filtering problem for multi-rate nonlinear systems with eavesdroppers and fading measurements, thereby improving the accuracy of the filtering performance.

The middle station positioning structure of the horizontal garbage compression station comprises a jacking frame and a compression plate, a compression bin is fixedly arranged at the top end of the jacking frame, a plurality of compression boxes are arranged in the compression bin, and a filter plate is fixedly arranged in the middle of the inner side of each compression box. A plurality of uniformly distributed filter holes are formed in the surface of the filter plate, electric telescopic rods are fixedly mounted on the two sides of the bottom end of the filter plate, fixing strips are fixedly connected to the movable ends of the two electric telescopic rods, and a plurality of connecting rods are fixedly arranged between the two fixing strips; according to the middle station positioning structure of the horizontal garbage compression station and the compression station of the middle station positioning structure, the two electric telescopic rods drive the two fixing strips to move upwards, then the multiple connecting rods and the multiple dredging rods are driven to move upwards, filtering holes in a filtering plate are dredged through the multiple dredging rods, and therefore the working efficiency of the horizontal garbage compression station is improved. The filter holes are prevented from being blocked, so that normal filtration of sewage is not influenced.

This invention discloses a distributed filtering method, system, and storage medium based on deep unfolding, relating to the field of sensor network technology. It aims to address the problems of poor adaptability to various noise levels, insufficient node estimation consistency, and high computational complexity in traditional methods. By establishing a network topology and system model, designing an information fusion filter structure, mapping the system iterative equations to a deep unfolded neural network, constructing a composite loss function containing estimation error and node divergence, and using the deep neural network for training to solve for the optimal gain, this invention is adaptable to noisy scenarios, achieving both high accuracy and consistency, with low computational complexity and strong robustness, making it suitable for fields such as environmental monitoring and dynamic target tracking.