6results about How to "Realize intelligent processing" patented technology

This invention provides a contract review assistance system, method, and storage medium. By constructing a four-layer architecture of "main agent - sub-agent - toolset - data resource layer," intelligent processing of contract review and related business scenarios is achieved. This architecture not only improves the flexibility and collaborative efficiency of task processing but also reduces reliance on manual intervention and error rates. Thus, it achieves an efficient, accurate, and structured review process through intelligent means, thereby improving the processing efficiency of contract review and related business scenarios. This solves the technical problem that existing contract review methods cannot fully meet the core review needs of the contract marketing and risk control domains, leading to low contract review efficiency.

The invention relates to a cross-border transportation control system and a control method thereof, and the system comprises a cross-border transportation management platform which is used for carrying out the transportation scheduling according to a transportation demand and a vehicle state, monitoring the transportation data and cargo information in a transportation process, and carrying out the data analysis and risk early warning; the customs inspection optimization system is used for analyzing the cargo information and pre-judging a customs inspection result; automatic customs inspection and unified management of data are carried out; the intelligent transportation decision-making system is used for planning a transportation path and allocating transportation resources according to cargo characteristics and transportation requirements; the cross-border logistics information integration and analysis system is used for collecting, processing and analyzing cross-border logistics data and deploying the data according to an analysis result; the cross-border transportation safety guarantee system is used for evaluating safety risks in the cargo transportation process and handling emergencies. Compared with the prior art, intelligent management and scheduling of the cross-border transportation process are realized, and the transportation flexibility and efficiency are improved.

This invention belongs to the field of data processing technology, specifically disclosing a safety assessment method for energy storage power stations based on real-time monitoring data. The method includes: performing multimodal feature analysis on voltage and environmental data sequences of the energy storage power station to obtain static voltage feature tensors and static environmental feature tensors; conducting collaborative analysis of the static voltage and environmental feature tensors to obtain an environmental voltage feature correlation group; performing entropy weight dynamic analysis on the environmental voltage feature correlation group to construct a multidimensional state vector; performing time-series deduction on the multidimensional state vector to obtain a safety situation evolution matrix; and performing residual analysis between the safety situation evolution matrix and real-time monitoring time-series data to obtain a comprehensive deviation index; and performing root cause assessment on the comprehensive deviation index to obtain comprehensive risk information. This invention enables more accurate, dynamic, and efficient safety assessment of energy storage power stations, allowing for comprehensive situation perception and rapid identification of risk root causes.

The invention relates to the technical field of metal plate bending processing, in particular to a real-time collision detection method based on self-adaptive bounding boxes and machine learning, which comprises the following steps: firstly, carrying out geometric feature analysis on each three-dimensional model, and calculating the fitting degree between the volume of the model and the volumes of different bounding boxes (sphere, AABB and OBB); and adaptively selecting and constructing an optimal bounding box with the highest space utilization rate. In a real-time detection stage, aiming at an object pair to be detected, 24-dimensional geometrical relationship feature vectors between bounding boxes of the object pair to be detected are extracted, and the 24-dimensional geometrical relationship feature vectors are input into a pre-trained lightweight multilayer perceptron model. The model outputs continuous collision probability values, and three-level intelligent classification (high-probability collision, critical state and safety) is performed according to a preset threshold value, so that scheduling of accurate detection resources is guided. According to the method, the package tightness of the bounding box can be improved from the source, and the boundary condition is accurately pre-judged by utilizing machine learning, so that the efficiency of real-time collision detection is improved while the detection precision is ensured.

The invention relates to the field of engineering cost, in particular to an engineering cost index analysis method based on reinforcement learning, which comprises the following steps of: identifying a table contour and a standard font contour in file image data, generating optimized image data, extracting a pointing keyword as a label, and forming a clustering set based on semantic similarity clustering; visual similarity analysis is carried out by reserving a table contour, a common label is set, a mapping association network of a keyword label and the common label is constructed, and intelligent processing of data uploaded by a user side is achieved. When the common label is successfully associated, a key data area is positioned according to area information, and local analysis is carried out; and if not, carrying out traversal analysis and semantic understanding on the complete table structure. According to the method, manual marking is replaced by automatic feature extraction, structured association of engineering knowledge is established, differentiation processing strategies are intelligently switched according to the matching degree, and the training data quality and the generalization ability of the model in a complex engineering scene are improved.

The invention discloses a self-adaptive network anomaly detection and policy-driven automatic recovery method and device, and relates to the technical field of network security and network management. The method comprises the following steps: collecting network multi-source indexes, and performing time sequence decomposition and pseudo data elimination to obtain preprocessed data; a weighted distance model is constructed to calculate an anomaly score, and adaptive anomaly detection is realized; after abnormity is determined, an abnormity propagation model is constructed based on network topology and node interaction strength, and an abnormal source node and a propagation path are determined; performing abnormal root cause analysis through graph structure similarity matching; according to the anomaly detection and root cause analysis result, constructing a multi-objective optimization model to generate a differential recovery strategy and executing the differential recovery strategy; the recovery effect is monitored and fed back, and the detection and strategy model is optimized. According to the method, a closed-loop self-healing process of detection-diagnosis-decision-execution-verification is formed through multi-dimensional feature modeling, anomaly propagation analysis and an optimization decision mechanism, and the accuracy and recovery efficiency of network anomaly detection are effectively improved.