6results about How to "Effectively filter out interference" patented technology

The application discloses a kind of based on neighborhood relative entropy multiscale subspace electric load anomaly detection method, including obtaining electric load data, constructs original data matrix and pre-processes;Using principal component analysis to the data matrix is extracted and is reconstructed, constructs multiscale subspace and generates reconstructed sample set;For each scale's subspace, calculate reconstruction error and outlier score;For reconstructed sample set, construct subspace neighborhood information system, and define neighborhood and neighborhood relationship in system;Calculate neighborhood relative entropy outlier score;Reconstruction error and neighborhood relative entropy outlier score are weighted and fused, obtain final fusion outlier score, and whether abnormal is determined according to threshold value.The application removes noise and redundant information in data effectively by multiscale principal component analysis, enhances the capture ability to global anomaly pattern, uses subspace neighborhood information system to mine the uncertainty of local neighborhood, improves the precision and robustness of anomaly detection under complex environment.

This invention discloses a method and system for geological hazard identification based on multimodal semi-supervised learning. The method includes the following steps: acquiring labeled and unlabeled orthophotos and elevation models of geological hazard areas; constructing teacher and student networks containing dual-branch encoders, extracting features from the orthophotos and elevation models respectively, and linearly fusing them, generating ensemble predictions using learnable dynamic weights; in semi-supervised training, calculating the uncertainty of the student network's prediction results in real time, and triggering an exponential moving average update of the teacher network using the student network parameters only when the decrease in uncertainty compared to the historical mean exceeds a preset evolution threshold; optimizing the student network parameters by combining supervised loss, confidence-weighted consistency loss, and student historical consistency loss; inputting the test data into the trained student network, and outputting fine-grained geological hazard segmentation results. This invention can achieve high-precision automatic identification of geological hazards under small sample conditions.

This invention discloses an intelligent environmental monitoring system for server racks, comprising: an edge perception layer deployed in the server rack for collecting environmental time-series data; a data governance layer including a data verification module, a compression storage module, and a feedback iteration module for anomaly detection, layered compression, and model updates; a collaborative judgment layer including a consensus module for cross-validating anomalous data using the PBFT consensus algorithm; an intelligent control layer including a control module for generating air conditioning control commands using a model predictive control algorithm; and a digital twin layer including a 3D model and a prediction module for predicting environmental change trends and visualizing them. This invention achieves a closed-loop end-to-end system encompassing data verification, lightweight storage, collaborative consensus, intelligent control, and digital twins, significantly reducing false alarm rates, optimizing storage space, improving control accuracy, and enabling predictive maintenance. It can be widely used for data center environmental monitoring.

This invention discloses a method and device for detecting material level in a closed silo, relating to the field of industrial automation detection and control technology. It employs a photosensitive level gauge to continuously acquire the original optical feedback signal of the solid material pile within the silo, receiving the light source sensor signal as the original optical feedback signal, forming signal frames in chronological order, with each frame corresponding to the light intensity value at a sampling time. Dynamic baseline tracking is performed on the acquired multiple signal frames, updating the baseline in real time using the weighted moving average of historical signal frames, and extracting the current optical signal trend component. The baseline update rate is automatically adjusted according to the degree of signal fluctuation. A multi-frame fluctuation suppression algorithm determines the deviation between the current signal frame and the baseline, dynamically filtering out instantaneous spikes caused by dust adhesion based on the statistical characteristics of the deviation. This invention significantly improves the detection accuracy, anti-interference capability, and operational reliability of automatic feeding systems for solid materials in kilns, and is suitable for industrial sites with high dust levels, variable light paths, and long operating cycles.

Embodiments of the present application disclose a vehicle data anomaly detection method, device and equipment, and a storage medium. The method comprises: acquiring kinematic parameters of a vehicle, and determining whether the vehicle is in an effective driving condition based on the kinematic parameters; in the case that the vehicle is in the effective driving condition, acquiring a current data change characteristic of a target collection index, and dynamically determining a judgment threshold corresponding to the target collection index based on the kinematic parameters; and determining whether the target collection index has a card stagnation anomaly based on the current data change characteristic, the judgment threshold, and a duration of the target collection index being in a low change state. Through pre-screening of the working condition and dynamic threshold adaptation, the present application realizes accurate identification of the card stagnation phenomenon of the collection index in the vehicle cloud data reporting link, and effectively reduces the false positive rate and the false negative rate.

The application relates to the technical field of building engineering monitoring, in particular to a building construction safety monitoring method and system based on sensors, which comprises the following steps: collecting real-time displacement monitoring data of each monitoring point of a building structure through a displacement sensor array, calculating the displacement value difference between each sensor node and its adjacent sensor node based on the real-time displacement monitoring data, and obtaining the displacement gradient value between the nodes; comparing the displacement gradient value between the nodes with a preset gradient threshold value, and activating the sensor array reconstruction mechanism when the displacement gradient value exceeds the preset gradient threshold value. In the monitoring process, the displacement gradient is used as a trigger signal to drive the idle nodes to automatically migrate to the high deformation gradient area, the dynamic matching of the sensor density and the deformation intensity is realized, the deformation capturing capability and the response speed of the monitoring area are improved, the corrected set is generated by fusing the strain and displacement data, the noise is filtered by using the time sequence correlation degree, the environmental interference is eliminated, and the stability of the monitoring data is ensured.