7results about How to "Has industrial application value" patented technology

The utility model discloses a kind of gas turbine waste heat boiler drum supports, including base, backing plate and the saddle for supporting drum, two pieces of polytetrafluoroethylene plate are provided between the backing plate and base, two pieces of polytetrafluoroethylene plate are fixed on backing plate and base respectively, two pieces of polytetrafluoroethylene plate can be freely relatively moved, so that the saddle on backing plate can be freely relatively moved with base along the axial direction of drum.The contact surface of the backing plate and base of the utility model gas turbine waste heat boiler drum support is polytetrafluoroethylene plate, and the friction coefficient is small, so that the saddle for supporting drum can be freely relatively moved along the axial direction of drum relative to base;When drum thermal state expands, the structure effectively releases stress in drum;Base is increased between backing plate and steel frame, avoid support to directly contact with uneven steel frame, improve the flatness of contact surface surface.

The application belongs to the technical field of neural network model training, and discloses a neural network model training method and system based on an improved random configuration algorithm, which comprises the following steps: under the condition of reducing the root mean square error of the neural network model output, screening suitable neurons as candidate hidden layer nodes through the inequality constraint condition of the random configuration algorithm; screening K neurons with the fastest training error reduction from the candidate hidden layer nodes, and selecting the neuron least related to the previous L-1 hidden layer nodes from the selected K neurons as the optimal hidden layer node; calculating the output weight through the least square method, and updating the structure of the random configuration network; and judging whether the network structure is completed by using the maximum allowable number of hidden layer nodes and the maximum allowable output error. The application can reduce the overall calculation amount of the algorithm while ensuring the prediction accuracy, and is suitable for application scenarios with high real-time requirements.

The invention discloses a particle board surface defect identification method and system based on a twin network and supervised comparative learning, and the method comprises the steps: (1) collecting a particle board surface image, carrying out the preprocessing, constructing a small sample data set containing various defects, and dividing the small sample data set into a training set and a test set; (2) constructing an improved twin supervised contrast network model, wherein the model comprises a feature extraction backbone network, an LDFPN and a classification contrast learning head; (3) training the twin supervised comparison network model by using the training set, and adopting joint optimization of cross entropy classification loss and supervised comparison loss during training; and (4) performing defect identification and classification on the shaving board surface image by using the trained twin supervision comparison network model. The method and the system aim at solving the problems of low particle board surface defect identification accuracy and weak model generalization ability under the condition of small samples, efficient and accurate automatic quality detection is realized, and the method and the system have the potential of real-time deployment on a production line.

The invention belongs to the field of preparation of high-reaction-activity isobutylene-based polymers, and relates to a high-reaction-activity isobutylene-based polymer and a preparation method thereof. The polymer comprises an isoolefin structural unit and a structural unit derived from a non-conjugated alkylene styrene compound, and the molar content of the structural unit derived from the non-conjugated alkylene styrene compound is 1-20% based on the total molar weight of all the structural units in the polymer; the polymer macromolecular chain contains external double bonds at the side group and the end group, and the average functionality of the external double bonds is not less than 1.0; the number-average molecular weight of the high-reaction-activity isobutene-based polymer is 500-6500 g / mol, the molecular weight distribution is narrow, and the distribution index can reach 1.3. The isobutene-based polymer with narrow molecular weight distribution and high reaction activity is directly prepared by a one-step method, side groups and end groups of the isobutene-based polymer contain external double-bond functional groups in a molecular chain, and the average external double-bond functionality is not less than 1.0.

This application discloses a method for synthesizing 2,6-difluorostyrene, belonging to the field of organic chemical synthesis technology, including the following steps: (1) 2,6-dichlorobenzonitrile undergoes a halogen exchange fluorination reaction with a fluorinating agent in a polar aprotic solvent to obtain 2,6-difluorobenzonitrile; (2) the obtained 2,6-difluorobenzonitrile undergoes an addition reaction with methyl magnesium halide in an inert solvent, followed by hydrolysis to obtain 2,6-difluoroacetophenone; (3) the obtained 2,6-difluoroacetophenone undergoes a carbonyl reduction reaction in a solvent to obtain 1-(2,6-difluorophenyl)ethanol; (4) the obtained 1-(2,6-difluorophenyl)ethanol undergoes a dehydration reaction in an alkane solvent under protic acid catalysis to obtain 2,6-difluorostyrene. This synthesis method has the advantages of inexpensive and readily available raw materials, good atom economy, simple reaction operation, high synthesis yield, and good product purity, making it suitable for industrial application.

The invention relates to an image feature extraction device based on a multi-scale convolutional network model and application thereof, and belongs to the technical field of artificial intelligence and computer vision, the device comprises the multi-scale convolutional network model based on dynamic weight self-adaption, the model comprises a self-adaption heterogeneous convolution kernel dynamic weighting deep convolution module, an image feature extraction module and a feature extraction module, the multi-scale convolution kernel fusion module is used for adaptively generating and fusing weights of multi-scale convolution kernels according to input image features; the dynamic Inception mixer is used for processing multiple groups of features after channel segmentation in parallel; the dynamic mixing block is used for extracting and fusing multi-scale features and channel interaction features through a dual-path residual structure; and the integrated network module based on the C2f architecture comprises a plurality of dynamic mixing blocks which are connected in series and is used for aggregating the multi-scale features and outputting final image feature representation. According to the method, multi-scale features and long-distance dependence in the image can be efficiently captured, the model performance is improved while the parameter quantity is reduced, and the method is suitable for various computer vision tasks such as target detection and image segmentation.

The present application relates to a kind of succinic acid preparation method of tartaric acid.The present application develops an innovative transfer hydrogenation reaction system and selectively generates succinic acid by hydrogenating tartaric acid to deoxidize, in using elemental iodine as transfer hydrogenation catalyst, methyl isobutyl ketone is used as hydrogen donor and solvent, and it can be reacted under normal pressure nitrogen using pressure-resistant tube, and tartaric acid is selectively converted into succinic acid by the process of transfer hydrogenation.85% succinic acid yield can be obtained.In addition, because the solubility of the generated product succinic acid in methyl isobutyl ketone is low, succinic acid will precipitate in solid form when cooled to room temperature, at this time the product can be conveniently separated by filtration.This method has high selectivity, uses cheap catalyst iodine, reaction condition is mild and product is easy to separate, easy to apply to industrialization, has great research and application value.