30results about How to "Increase the level of influence" patented technology

The invention provides an intelligent detection method based on a graph neural network. The intelligent detection method comprises the following steps of collecting data, preprocessing the data, building a network model, carrying out pre-training and transfer learning, carrying out predicting and performing casual inspection verification to perfect a whole prediction system. Compared with manual detection, the method has the advantages that the component detection accuracy and efficiency are improved, the interference of human factors on detection is reduced, and the labor cost and the detection cost are reduced. Compared with a traditional machine learning method, the method has the advantages that the graph neural network does not require that the composition form of the data must have agood spatial relationship, that is to say, the graph neural network has a neatly arranged matrix form, and the feature that the graph neural network can accept unstructured input significantly improves the expression ability of the model. Compared with a convolutional neural network method, the graph neural network can better learn the logic relationship of each element, so that the generalization ability of the model is improved. In the learning process of the network, each node is responsible for spreading own information and integrating information of neighbor nodes, so that the logic normal form of data is learned and mastered.

The invention discloses a no-Snapback silicon controlled rectifier type ESD protection structure and a realization method thereof. The structure comprises: a semiconductor substrate (80); and an N-well (60) and a P-well (70) that are generated on the semiconductor substrate. A high-concentration P-doped layer (20) and a high-concentration N-doped layer (28) are arranged above the N-well (60). Thehigh-concentration P-doped layer (20), the N-well (60) and the P-well (70) form an equivalent PNP triode structure. A high-concentration N-doped layer (24) and a high-concentration P-doped layer (26)are arranged above the P-well (70). The N-well (60), the substrate (80)/the P-well (70) and the high-concentration N-doped layer (24) form an equivalent NPN triode structure. A high-concentration N-doped layer (22) is arranged above a boundary between the N-well (60) and the P-well (70). A portion of the N-well (60) is arranged between the high-concentration P-doped layer (20) and the high-concentration N-doped layer (28), and the distance between the high-concentration P-doped layer (20) and the high-concentration N-doped layer (28) is S. A portion of the N-well (60) is arranged between the high-concentration N-doped layer (28) and the high-concentration N-doped layer (22).

The invention discloses a dangerous chemical road transportation drainage basin water environment risk assessment method, which adopts an environment sensitive receptor influence derivation method, takes an environment sensitive receptor as an assessment basis, screens an environment risk road section according to the influence degree of dangerous chemical leakage to be transported on the environment sensitive receptor, and grades the environment risk road section; identifies a plurality of environmental risk road sections according to the drafted hazardous chemical substance transportation route, and then performs the environmental risk assessment and environmental risk grade division one by one on the road sections to serve as an auxiliary decision basis for finally determining a hazardous chemical substance transportation route scheme. The invention further discloses a dangerous chemical road transportation drainage basin water environment risk assessment system. An environment riskassessment module identifies an environment risk receptor around the leakage path, an environment risk source and environment risk substance information; the risk source identified by the environmentrisk identification sub-module is evaluated on the basis of an environment risk source evaluation method, the environment risk level of the risk source is evaluated, and the decision-making module isassisted to realize the prediction simulation of the sudden environment event.

The embodiment of the invention discloses a fingerprint recognition method and a related device. The method comprises: collecting a first fingerprint image of a user finger through a fingerprint module; converting the first fingerprint image into a second fingerprint image, wherein each pixel point in the second fingerprint image is determined according to a plurality of pixel points in the firstfingerprint image; generating a plurality of fingerprint feature sets of the user finger according to the second fingerprint image, each fingerprint feature set comprising a fingerprint feature point,an associated texture feature of the fingerprint feature point, a feature point matching weight and a texture matching weight, the associated texture feature comprising description information for reflecting a contour of a line where the fingerprint feature point is located; performing fingerprint comparison operation on the plurality of fingerprint feature sets to obtain a comparison result of each fingerprint feature set; and determining a fingerprint identification result according to a comparison result of the plurality of fingerprint feature sets. According to the embodiment of the invention, the false rejection rate FRR of fingerprint identification is reduced, and the fingerprint identification accuracy of the electronic device is improved.

The invention discloses a geographic detection and PCA based method for screening a key influential factor of PM2.5 concentration distribution. The method comprises that the positions of PM2.5 monitoring stations are obtained, and buffer areas are established; influential factor data in an area to be researched is obtained and made discrete; corresponding influential factor data of each PM2.5 monitoring station is obtained from the discrete influential factor data; and the contribution degree and relevance of the discrete influential factor data are analyzed by a geographic detector model. The problem that the prior art lacks an analysis method for comprehensively considering relation between variables and PM2.5 concentration distribution can be solved, the accuracy is high, analysis is comprehensive, and the application range is wide.

The invention provides a no-snapback silicon controlled rectifier type ESD protection structure and an implementation method thereof. The structure comprises a semiconductor substrate (80) and an N well (60) and a P well (70) produced from the semiconductor substrate; a high-concentration P-type doping (20) and a high-concentration N-type doping (28) are located on the upper portion of the N well(60); the high-concentration P-type doping (20), the N well (60) and the P well (70) form an equivalent PNP triode structure; a high-concentration N-type doping (24) and a high-concentration P-type doping (26) are located on the upper portion of the P well (70); the N well (60), the base body (80)/P well (70) and the high-concentration N-type doping (24) form an equipment NPN triode structure; a high-concentration P-type doping (22) is located above the dividing boundary of the N well (60) and the P well (70); the high-concentration P-type doping (20) is isolated from the high-concentration N-type doping (28) via a shallow trench isolation layer (10) with a width of S; a part of the N well (60) is between the high-concentration N-type doping (28) and the high-concentration P-type doping (22).

The invention discloses a risk assessment method for a sudden drainage basin water environment of a hazardous chemical production enterprise, and the method employs an environmental sensitive receptorinfluence derivation method, takes an environmental sensitive receptor as an assessment basis, and carries out the environmental risk assessment according to the influence degree of the leakage amount of substances stored or produced in an enterprise factory on the environmental sensitive receptor. And for the region with the environmental risk, determining the environmental risk level of the region in combination with the level of the environmental sensitive receptor. And further according to the provided dangerous chemical storage planning line, taking the dangerous chemical storage planning line as an auxiliary decision-making basis for finally determining a dangerous chemical storage scheme. The invention further discloses a risk assessment system for the sudden drainage basin water environment of the hazardous chemical substance production enterprise, the risk assessment system helps the enterprise to carry out daily environment risk management, early warning, environment emergency disposal and other work, and the environment risk management level is improved. And meanwhile, the prediction simulation of an emergency environment event is realized, and the environmental emergency disposal level is improved.

The invention discloses a dangerous chemical substance road transportation ecological environment risk assessment method, which comprises the following steps: (1) setting a risk assessment system based on a network, and (2) setting an environment risk assessment management module; (3) setting an environmental risk assessment sub-module; and (4) setting an environmental risk level evaluation modulefor determining the environmental risk level of the road section with the ecological environmental risk by combining the level of the ecological environmental sensitive receptor. The risk assessmentsystem comprises a basic information management module, an environment risk assessment management module, an environment risk assessment sub-module and an environment risk level assessment module which are mutually connected and run based on the network, and the level of the ecological environment sensitive receptor is combined with the road section with the ecological environment risk. Accordingto the invention, an ecological environment sensitive receptor influence derivation method is adopted to realize prediction simulation of an emergency environment event, which is helpful for improvingthe environment emergency disposal level.

The invention provides an information recommendation method and device, equipment and a storage medium. The method comprises the steps of receiving a recommendation request, wherein the recommendation request comprises an identifier of a user; determining a candidate object; according to the identifier of the user and a preset corresponding relationship, determining a user feature vector corresponding to the user, the corresponding relationship being used for representing a mapping relationship between the user and the user feature vector, and the user feature vector being obtained by training vectorized representation of the user in the user data; the vector is used for representing the interest degree of the user on the object and the influence degree of the interest of the social friends of the user on the interest of the user; according to a user feature vector and a feature vector of the candidate object, the recommendation degree of the candidate object to the user is determined, and the feature vector of the candidate object is obtained by training vectorized representation of the object in the user data and is used for representing a vector of a user group of the object; and sorting the candidate objects according to the recommendation degree to obtain a recommendation result, and returning the recommendation result to the terminal equipment of the user.

The embodiment of the invention provides a node similarity calculation method, device and equipment and a computer readable storage medium. The method comprises the steps of obtaining a bipartite graph, wherein the bipartite graph comprises nodes of a first type and nodes of a second type; according to the bipartite graph, determining a first set and a second set, connecting each node in the firstset with the first node of the first type, and connecting each node in the second set with the second node of the first type; according to the number of the same nodes in the first set and the secondset, the first weight value of the connection line between the first node and each node in the first set, the second weight value of the connection line between the second node and each node in the second set, and the similarity between each node in the first set and each node in the second set, determining the similarity between the first node and the second node. According to the embodiment ofthe invention, the accuracy of calculating the similarity between the low-weight distributed nodes is improved.

The invention discloses a driving motor control system for a vehicle, which comprises a parameter input module, N controllers preset based on N different working conditions, a stacking module, a pulsewidth modulation module and a working condition recognition module, wherein the parameter input module is used for sending the difference between i<d><*> and id and the difference between i<q><*> andiq to each controller; each controller is used for outputting a corresponding voltage control amount according to input signals of each controller; the stacking module is used for calculating ud anduq through first equations, wherein the first equations are as described in the specifications, and wi is the weight corresponding to ith controller; the pulse width modulation module is used for generating corresponding pulse width modulation signals according to ud and uq for motor control; and the working condition recognition module is used for recognizing a controller closest to the actual working condition of the current vehicle in the N controllers according to a preset recognition cycle and improving the weight corresponding to the recognized controller. When the scheme disclosed in the invention is applied, the vehicle can reach high efficiency and high performance under different working conditions. The invention also discloses a vehicle and a driving motor control method thereofwith the corresponding effects.

The invention discloses a preparation process of corn edible oil, which comprises: 1, mixing corn germ powder and citric acid according to a certain weight part ratio, extracting, and evaporating to remove a solvent to obtain corn crude oil; 2, performing standing and precipitating the crude oil for 40-60 minutes, filtering by using a filter, and deodorizing; 3, roasting the whole corn grains at the constant temperature of 50-60 DEG C through a stirring tank, and putting the roasted corn grains into a squeezer to be squeezed to obtain oil; and 4, washing the white sesame seeds and the perilla seeds with warm water, drying in the air, mixing and stirring with all the raw materials, heating to 100-115 DEG C, reacting for 1-5 minutes, standing, and drying to 35-40 DEG C. The method has the beneficial effects that the corn germ powder and the whole corn grains are used for branch type crude oil squeezing, the content and the quality of the crude oil in the initial step can be improved, so that the process treatment delicacy is improved, and the required reaction conditions in the whole preparation process become milder.

The invention discloses an unbalanced node classification method based on a minority class weighted graph neural network, and the method comprises the steps: 1, calculating a node membership value based on adjacency information for input graph structure data, so as to obtain the weighted feature information of a node in an embedded space; 2, executing a data oversampling operation in the embedding space to generate a new node; 3, cost sensitive learning is introduced in the training process of the edge predictor, and the trained edge predictor is used for obtaining adjacency information of a new node; and 4, constructing balanced graph structure data in combination with the features of the new nodes and the existing nodes and the adjacency information, performing node classification by using a graph neural network, and using a Gumbel distribution function as an activation function. According to the method, the neighbor aggregation of minority class nodes is enhanced, the bias to majority class nodes in the edge generation process is reduced, the convergence speed of the graph neural network model is improved, and the unbalanced node classification effect based on the graph neural network is remarkably improved.