144results about "Forecasting" patented technology

Embodiments are described for a system and method of controlling access to information in an organization by defining a hierarchical organizational structure of boxes, and security configuration comprising user records, security roles, rules to map users to boxes, and rules to grant roles to users via mapped boxes. Access control is applied in the context of a defined organizational structure using the effective set of access control policies computed in real time per each data access request from any given user.

The invention discloses a taxi passenger-searching path recommendation method based on information entropy. The steps include: mining a taxi track, and extracting passenger carrying point data therefrom; mining and extracting passenger-searching points representing passenger gathering places from the passenger carrying points obtained by extraction; measuring attribute values of factors according to which a taxi selects a passenger searching position in a passenger searching process; according to the current position and current time of the taxi, obtaining passenger searching points in a designated range, and building a passenger searching point selection model which includes a set of passenger searching points where the taxi can arrive departing from the current position at the current time and a decision matrix; solving the passenger searching point selection model based on information entropy, and obtaining a designated number of passenger searching points with optimal synthesized attributes; and using the passenger searching point obtained at the previous time as a reference point, repeatedly building and solving the models for a designated number of times, and generating a passenger searching path according to an obtaining level of optimal passenger searching points and recommending the path. The taxi passenger-searching path recommendation method based on information entropy has the advantages of being capable of bring good benefits to a taxi driver and being high in recommendation accuracy.

The system and method of the present invention automatically extracts the top k recommendations of objects, such as topics, items, products, books, movies, food, drinks, etc., from a local probabilistic recommendation system. Unlike prior systems, the present invention accomplishes the extraction of the top k recommendations of objects without examining a probability for every object that can be recommended. Further, the system and method of the present invention is capable of being implemented using probabilistic recommendation systems based on any conventional type of probabilistic distribution or machine learning technique, including, for example, decision trees and Bayesian networks.

The invention discloses a weather index and weighted LS-SVM-based power system short-term load prediction method. The method comprises the following steps of S1, obtaining a sample of original data; S2, according to the original data, calculating a comprehensive weather index; S3, performing data preprocessing on date type data and the comprehensive weather index; S4, according to an obtained dimensionless load characteristic quantity, performing gray correlation analysis between the dimensionless load characteristic quantity and a power system load, and calculating a characteristic quantity weight through a correlation degree obtained by the gray correlation analysis; and S5, building a weather index and weighted LS-SVM-based power system short-term load prediction model, performing parameter optimization by adopting a fruit fry optimization algorithm, and obtaining power system load prediction data of a to-be-predicted day through model output. The method has very good global optimization performance and few adjustment parameters, difficultly falls into local minimum and can effectively improve the power system short-term load prediction precision.

The embodiment of the invention provides a construction method and device for a double-layer path of a vehicle-mounted unmanned aerial vehicle. The method comprises the steps that cluster grouping isperformed according to a minimum distance of a task target point space of the unmanned aerial vehicle, and a Hamilton path conforming to unmanned aerial vehicle cruising ability constraint of each group is constructed; and the Hamilton paths corresponding to all the groups are merged and adjusted through a connection algorithm, and a loop starting from a base point through a starting point and anending point of each Hamilton path back to the base point is constructed. Through the method, a double-layer path planning scheme of the vehicle-mounted unmanned aerial vehicle during cooperative collection of target information is quickly established, it is guaranteed that the vehicle-mounted unmanned aerial vehicle travels by one loop in a path network within a planning area, the unmanned aerialvehicle completes collection of information at all task target points, and working efficiency is greatly improved.

The invention discloses a public bicycle station supply and demand prediction method based on a Markov chain. A balance stable equation relevant to station importance is established by constructing a transfer probability matrix of borrowed and returned vehicle by using the bicycle lease data of a public bicycle lease station terminal in order to predict the daily borrowing and returning demands of a station. The method has the positive effects that the classical method of Markov chain on probability statistics is applied in combination with the practical problem of a public bicycle lease station, and a practicable vehicle borrowing and returning supply and demand prediction method is provided, so that theoretical guidance is provided for the specific pile construction problem and balance scheduling problem of the public bicycle station. The method has a good industrial application prospect.

A microgrid photovoltaic power station generating capacity combined forecasting method belongs to the new energy power generation output forecasting field. The method comprises the following steps: 1) taking generating capacity data of a photovoltaic power station in a microgrid as input/output data of a forecasting model; 2) carrying out generating capacity forecasting through wavelet transform and a BP neural network; 3) carrying out generating capacity forecasting by utilizing a particle swarm algorithm to optimize a support vector machine; 4) obtaining model test weights by utilizing a variance-covariance combination method; and 5) calculating combined forecasting result according to the weights and model output corresponding to the new input. Compared with an existing forecasting method, the forecasting method starts from a new angle, avoids considering complex external factors and adopts a combination method, and has the advantages of being good in real-time performance, high in accuracy and high in adaptability and the like.

The embodiment of the application discloses a method and a device for determining a risk control strategy based on a prediction model. The method comprises the following steps: acquiring first user data and determining a target service associated with the first user data; according to the target business type of the target business, the target risk decision-making rule model associated with the target business is determined. A target risk decision rule corresponding to the first user data is determine based on the target risk decision rule model, a target risk decision rule model is obtained based on the sample user data training, and the sample user data comprises a first sample user data corresponding to a first risk decision rule and a second sample user data corresponding to a second risk decision rule; the target risk decision rule model comprises a first sample user data corresponding to a first risk decision rule and a second sample user data corresponding to a second risk decision rule. According to the target risk decision rules, the target risk control strategy is determined. By adopting the embodiment of the present application, the relevance between the risk control strategy determined based on machine learning and the user service data can be enhanced, the reliability of the risk control strategy can be improved, and the applicability is stronger.

The invention relates to the technical field of big data, is applied to the financial industry, and provides an enterprise potential risk early warning method and device, computer equipment and a storage medium. The method comprises the following steps: obtaining a sample; obtaining an enterprise association map and extracting a node association relationship; obtaining a risk parameter tag carriedby a propagation starting node in the enterprise association map, and obtaining a propagation path of the risk parameter tag according to the association relationship between the risk parameter tag and the node, obtaining a propagation coefficient among the nodes in the propagation path, and carrying out tag propagation processing on the risk parameter tag according to the propagation path and the propagation coefficient to obtain potential risk early warning information of the node. According to the scheme, an enterprise association map and label propagation processing are carried out; According to the technical scheme, the propagation path and the propagation coefficient are determined by utilizing the incidence relation between the corresponding nodes of the enterprise, effective propagation of the risk parameter tags is carried out, enterprise potential risk early warning is obtained through analysis more quickly and accurately, and compared with a manual analysis mode, the enterprise potential risk analysis efficiency and early warning reliability are improved.

Methods, apparatuses, and computer program products for provisioning aggregate computational workloads and air conditioning unit configurations to optimize utility of air conditioning units and processing resources within a data center are provided. Embodiments include for each air conditioning unit within the data center, determining a thermal zone generated by the air conditioning unit; for a given aggregate computational workload, identifying a plurality of computational workload configurations, each computational workload configuration indicating spatial assignments of the aggregate computational workload among a plurality of processing resources within the data center; for each computational workload configuration, calculating a total minimum energy consumption of the air conditioning units and the processing resources; and selecting the computational workload configuration with the lowest total minimum energy consumption and the determined lowest power air conditioning unit configuration corresponding with the selected computational workload configuration.

The embodiment of the invention discloses a production order production scheduling management method and production scheduling optimizer system. The method comprises the steps that unfinished orders are inquired; rearrangement plan settings are submitted; and order scheduling is performed. The optimal configuration of the production orders and production lines can be realized through production order production scheduling management so that scientific management of garment production can be realized, and the management efficiency and the production efficiency can be enhanced.

The invention discloses an order distribution scheduling and mixed product ordering system and method. The system comprises an input module, a cloud storage module, a processing module and an output module. The input module, the processing module and the output module are connected with the cloud storage module. The method comprises steps of determining production line available time according to real-time demands of orders, real-time demands of mixed products of scheduled orders in a former planned period and total demands of quasi-scheduled mixed products in a current planned period; distributing orders of different mixed products and delivery times to different parallel processing lines; and comprehensively considering setting time of products on each machine table to generate a mixed order-product production scheduling plan of each production line, maximizing the probability of completing products on each processing line in time in each plan period, and timely updating production programs of a numerical control machine tool. The objects of properly distributing different client orders to parallel production lines by batch according to differences in delivery times and items and further generating a production scheduling plan of each production line are achieved.

The invention discloses a single-vehicle multi-site distribution line processing method. The method includes: planning a line of a single site to multiple distribution sites by employing a saving algorithm based on map data, and obtaining planned initial distribution line parameters, wherein the multiple distribution sites comprise two or more distribution sites in a logistics distribution process, and the single site refers to a distribution origin or a distribution terminal in the logistics distribution process; and performing optimization processing on the initial distribution line parameters by employing a genetic algorithm, and obtaining practical distribution line parameters. According to the method, through real map road network data, the calculation of the sequence of the distribution sites is more reasonable based on the calculation of the optimal path of distribution sites in pairs through the combination of the saving algorithm and the genetic algorithm, the distribution efficiency of logistics enterprises can be improved, and the distribution cost can be effectively lowered.

The invention discloses a crowd behavior simulation method and a crowd behavior simulation system guided by a safety evacuation mark. The method comprises the following steps of: creating a three-dimensional building model of a scene to be simulated; determining the quantity and the position of the exit corresponding to the scene to be simulated in the three-dimensional building model; marking thelocation of safety evacuation signs corresponding to the scene to be simulated in the three-dimensional building model; according to the number of exits, the location of exits and the location of safety evacuation signs, the road markers in the three-dimensional building model being set in advance, and then PRM algorithm being used to generate all the available evacuation paths, and the paths inthe planning direction of safety exit being marked as one-way paths, paths beyond the planning direction of the safe exit being marked as one-way paths to the outbound exit; importing characteristic data of initializing evacuation crowd; a Dijkstra algorithm being used to find the shortest escape path from the one-way path for the evacuated individuals; in the process of crowd evacuation, an RVO obstacle avoidance strategy being used to simulate crowd evacuation.

The invention relates to the field of intelligent maintenance of semiconductor manufacturing, and discloses a predictive maintenance and health management method of a semiconductor manufacturing device. A modeling strategy module, a model training module and a fault prediction module are contained, the modeling strategy module determines a key component and a modeling strategy of the device, the model training module only uses health state data to train the model of the key component, and the fault prediction module updates the model according to real-time data and predicts a key fault type inreal time; and the modeling strategy module outputs the key component that needs predictive maintenance, and the key failure mode that should be used as a predictive maintenance target to serve as the inputs of the model training module, the model training module trains the model of the key fault of the selected key component, and finally the fault prediction module establishes a model growth system to perform fault prediction.

The invention provides a method for predicting social attention hotspots in network public opinion events based on group analysis, which can predict the network public opinion events which may be developed into the social attention hotspots, provide decision support for processing the public opinion events, collect related data of the public opinion events, construct a public opinion event set, obtainpublic opinion event relation subgraphs and social attention indexes in different time windows, carry out graph representation learning on data in the public opinion event relation sub-graphs in different time windows through a graph embedding technology to obtain feature vectors of corresponding event relations, and enable the feature vectors of the event relations and corresponding social attention indexes to form a two-tuple, construct a time sequence prediction model based on a bidirectional long-short-term memory neural network, input the two-tuples into the model for training iteration until the model converges, input new public opinion events into the model to obtain social attention indexes of the new public opinion events in future time, and select the public opinion event with the maximum social attention index as a predicted social attention hotspot.

The invention discloses a high-speed railway subgrade settlement prediction system based on GEP. Firstly, the filling soil mass of a high-speed railway has different engineering properties due to different soil foundations, and a subgrade settlement prediction problem is researched by selecting typical road section data; secondly, according to the principle of gene expression programming, a high-speed railway subgrade settlement predication program based on the gene expression programming is programmed by using C++; and then, detection data of railway subgrade settlement are used as training data to be imported into the high-speed railway subgrade settlement predication program based on the gene expression programming, and a railway subgrade settlement predication function model based on the gene expression programming is obtained through program operation. Settlement prediction is carried out on a certain high-sped railway subgrade monitoring point by using the model and a GM(1,1) grey prediction function model, a prediction result is obtained, and the prediction accuracy of the two models is comprehensively analyzed, so that excellent prediction ability of the models is verified.

The invention relates to the field of artificial intelligence, discloses a traffic jam prediction method, device and equipment and a storage medium. The method is used for solving the technical problem that a traffic jam condition cannot be predicted in an existing urban intelligent traffic system. The method comprises the following steps: according to a traffic jam prediction request, obtaining traffic flow data at each traffic gate to obtain a traffic flow data set; analyzing the parameters of the traffic checkpoints in the traffic flow data set by using a big data analysis technology to obtain node parameters; calculating the relationship among the traffic checkpoints to obtain edge parameters; and generating a traffic map in a preset time period based on the node parameters and the edge parameters through a pre-constructed road network knowledge graph, and inputting the traffic map into a traffic prediction model pre-established based on a graph convolutional neural network for road condition prediction processing to obtain a predicted value of the congestion degree of each traffic gate corresponding to the traffic congestion prediction request. In addition, the invention also relates to a block chain technology, and the related information of the traffic flow data can be stored in a block chain.

The invention provides a method for optimizing a cutting path of a spacer frame type integral structural member, belongs to the technical field of design and manufacturing of modern structures, and solves the problem in optimization of the cutting path in a processing process of the spacer frame type integral structural member. The method comprises the following steps of (1) establishing a finite element model for a single spacer frame of the spacer frame type integral structural member, and loading initial internal stress; (2) performing dynamic cutting simulation on the single spacer frame to obtain an optimized cutting path; (3) establishing a finite element model for the spacer frame type integral structural member, and loading initial internal stress; and (4) performing dynamic cutting simulation on the whole spacer frame to obtain an optimal cutting path. According to the method for optimizing the cutting path by adopting dynamic cutting simulation, analysis and prediction can be performed in advance before parts are actually processed, and the cutting path can be optimized, so that a corresponding processing policy is adjusted, processing deformation is effectively controlled, the production cycle of the parts is shortened, and the production cost is reduced.

The invention relates to a short-term power load forecasting method which is characterized by comprising the steps of: selecting a similar day set; constructing a RBF (Radial Basis Function) neural network forecasting model optimized by a bat algorithm; forecasting a power load on a forecasting day by utilizing the RBF neural network forecasting model optimized by the bat algorithm; and the like.According to the invention, on the basis of conventional gray correlation analysis, a similar day with a higher similarity is selected by adopting a distance similarity and shape proximity correlatedcomprehensive gray correlation degree, the defect that a conventional gray correlation analysis method only considers a geometric similarity degree among data sequences, but ignores a number proximitydegree when selecting the similar day is made up, and forecasting accuracy is improved; a weight of a RBF neural network is optimized by utilizing the bat algorithm, so that the defect that the RBF neural network is easy to fall into local optimization can be overcome, a convergence rate of the integral network is improved and computation efficiency of the integral network is improved; and the short-term power load forecasting method has the advantages of scientificity, reasonability, high applicability, good effect and the like.

The invention discloses an intelligent logistics management method for an allocation center. The departure cargo collection process is completed by automatically scheduling vehicles. Departure cargo quantity prediction and loading execution are performed on departure cargos. Arrival cargo quantity prediction and unloading execution are performed on arrival cargos. Delivery quantity prediction andloading execution are carried out on cargos needing to be delivered. Different operation instructions of the whole process are snet to related personnel. Resources of the allocation center, the vehicles and the business outlets are integrated. Real-time analysis, unified management and unified scheduling are carried out; vehicles entering and exiting a dispatching center are intelligently dispatched through an intelligent reservation platform, and group loading and unloading are intelligently dispatched. The logistics operation efficiency of cargo circulation, personnel operation, vehicle loading, site use and the like of the allocation center and the area covered by the allocation center is improved. The customer requirements can be met, the customer satisfaction is improved, the operation cost of the allocation center is reduced, and the core competitiveness of enterprises is improved.

The invention relates to an interest point check-in prediction method fusing deep learning with a factorization machine and belongs to the field of location check-in prediction. The method comprises the following steps: S1, acquiring check-in data of a user; S2, performing embedding processing on input discrete data; S3, performing sparse elimination processing on the discrete data, and learning implicit second order relations among the data; S4, learning addition of continuous characteristics into a full connection hidden layer, and selecting an appropriate excitation function; S5, inputtinga result obtained by processing discrete characteristics and a result obtained by processing the continuous characteristics and adding the results as an input of a hidden layer h1; S6, enabling an output l1 of the hidden layer h1 to pass a first-order linear and characteristic interaction structure and adding as an input of a hidden layer h2; and S7, receiving an input by a hidden layer h3 from outputs l1 and l2 of the hidden layers h1 and h2, adding a shortcut structure at the same time for guaranteeing gradient stability during parameter learning, determining the best model structure, and finally outputting a prediction result. The method provided by the invention fully excavates and learns check-in rules and predicts interest point check-in problems by analyzing check-in information ofthe user.

The invention discloses an iterated local search algorithm by employing an improved perturbation mode for solving a job-shop scheduling problem. According to the algorithm, two neighborhood structures and an elite pool strategy are introduced and the perturbation mode of iterated local search is improved, so that the local search performance of the algorithm is improved, and the problems that the algorithm runs into local optimum during searching and the search range is reduced due to the fact that only a current optimal solution is perturbed during searching are solved. The implementation process of the algorithm comprises the steps of generating an initial solution; setting the capacity of an elite pool; executing neighborhood searching to randomly generate two neighborhood structures; and executing perturbation of the elite pool strategy. Compared with the prior art, the iterated local search algorithm has the advantages that the actual job-shop scheduling problem can be better solved, and the iterated local search algorithm has the characteristics of high practicability, high executable ability and the like.

The invention provides a method for optimizing a power grid planning scheme based on a comprehensive cost model. According to the method, the comprehensive cost model, which adopts the planning target age limit as a boundary and considers the annual cost and the current value of a planned scheme at the end of the year, is created. The calculating range of the model is the grid equipment (substations and transmission lines) which produces variable amount of costs in the scheme, wherein economic life of old equipment is needed to be forecasted to determine whether needing to update and when updating within the planning period. Meanwhile, fault rate forecast of new and old equipment within the planning period is trained and output through a least square support vector machine algorithm. In addition, when the comprehensive cost model is applied to comparing planning schemes, in order to reduce the influence on error of economic variables, an interval method is adopted. Numerical examples show that economic optimization of the power grid planning scheme is effectively realized by combining the comprehensive cost model with the interval method, and the comprehensive cost model is suitable for planning target age limits in various lengths.