Abstract: An image defogging method based on dark channel prior is provided, including: acquiring a dark channel image and a light channel image of a haze weather image according to a haze weather image; selecting pixel values in the light channel image and calculating an atmospheric light value A; obtaining a transmittance image based on the light channel image and the atmospheric light value A; obtaining a restored image of defogging based on the transmittance image and the atmospheric light value A. By using the pixel values of the light channel to distinguish whether pixel points comply with the dark channel prior, calculating the transmittance of image areas that comply with the dark channel prior and those that do not, so as to avoid the situation where the image defogging effect is not good due to inaccurate transmittance estimation, making the defogging of images containing high brightness areas more clear.

Abstract: An audio visual haptic signal reconstruction method includes first utilizing a large-scale audio-visual database stored in a central cloud to learn knowledge, and transferring same to an edge node; then combining, by means of the edge node, a received audio-visual signal with knowledge in the central cloud, and fully mining semantic correlation and consistency between modals; and finally fusing the semantic features of the obtained audio and video signals and inputting the semantic features to a haptic generation network, thereby realizing the reconstruction of the haptic signal. The method effectively solves the problems that the number of audio and video signals of a multi-modal dataset is insufficient, and semantic tags cannot be added to all the audio-visual signals in a training dataset by means of manual annotation. Also, the semantic association between heterogeneous data of different modals are better mined, and the heterogeneity gap between modals are eliminated.

Abstract: Disclosed is a multi-scale fusion defogging method based on a stacked hourglass network, including inputting a foggy image into a preset image defogging network; and outputting a fogless image after the foggy image is processed by the image defogging network. The image defogging network includes a 7×7 convolutional layer, a stacked hourglass module, a feature fusion, a multi-scale jump connection module, a 1×1 convolutional layer, a 3×3 convolutional layer, a hierarchical attention distillation module, the 3×3 convolutional layer and the 1×1 convolutional layer connected sequentially.

Abstract: A source-grid-load-storage networked collaborative frequency control method is disclosed, which comprises acquiring total active power ?P to be regulated during a secondary frequency regulation process of a power grid; performing frequency regulation by source-grid-load-storage of the power distribution system, allocating power regulation capacities, and determining whether the desired total active power is met after the frequency regulation; if the desired total active power is met, determining whether power of power generation units is out of limit; if the power of the power generation units is not out of limit, keeping active power of the power distribution system in balance to complete frequency regulation of the power grid; if the power of the power generation units is out of limit, correcting the power regulation capacities of the power generation units of the source-grid-load-storage and compensating a power difference to keep the active power of the power distribution system in balance.

Abstract: A common control channel-free communication method and system featuring integrated autonomous link establishment and frequency conversion, belonging to the technical field of wireless communication is provided. The method comprises: autonomously establishing, by a caller communication node a two-way communication link with a responder communication node through a calling and responding mechanism; and monitoring, by the caller communication node, operating frequency points through a real-time interference detection mechanism, and autonomously updating transmitted signal feature parameters of the caller communication node at certain frequency conversion slots, so that the responder communication node can update transmitted signal feature parameters in the responder communication node after receiving the updated transmitted signal feature parameters.

Abstract: A method for small object detection in drone scene based on deep learning is provided, which includes: inputting images captured by a drone into a pre-trained generator based on an Unet network structure to output normal-light images; inputting the normal-light images into a object detection backbone network to output a plurality of multidimensional matrix feature maps, wherein the object detection backbone network integrates a channel attention mechanism and a spatial attention mechanism based on convolutional block Self-Block, and a 7*7 large convolutional kernel is used; inputting the plurality of multidimensional matrix feature maps into a BiFPN-S module of a feature pyramid for feature fusion, so as to output a plurality of corresponding feature maps for predicting objects of different sizes.

Abstract: A networked control method for primary frequency regulation of a new energy power station based on a source-grid-load-storage networked cloud decision control system platform comprises: determining, according to historical operating data of a new energy power station, primary frequency regulation predictive values of power generation units of the new energy power station; determining, according to the primary frequency regulation predictive values, optimal control sequences of inverters of the power generation units at different times based on a pre-established inverter active power model, wherein the optimal control sequences comprise multiple control quantities of active power of the inverters; marking the optimal control sequences at the different times with time scales, sending the optimal control sequences to executing devices of the power generation units, receiving the optimal control sequences, and determining whether to store or not store the optimal control sequences; and determining the control qua

Abstract: The present disclosure discloses an image segmentation method combined with superpixel and multi-scale hierarchical feature recognition. This method is based on a convolutional neural network model taking multi-scale hierarchical features extracted from a Gaussian pyramid of an image as a recognition basis, and then being connected with a multilayer perceptron to achieve the recognition of each pixel in the image, moreover, this method is used tier performing superpixel segmentation on the image and is combined with a method for improving superpxiel in combination with LBP texture features to segment an original image so that an obtained superpixel block is more fitted to edges of targets, then, the original image is merged according to a mean value of a color, and finally, recognition of each target in the image is achieved.

Abstract: The present disclosure discloses a multi-task joint computing unloading and resource allocation method based on D2D (Device-to-Device) communication. For solving the problems that computing resources of a local terminal cannot complete all computing tasks on time when there are multiple computing tasks with delay requirements on a local terminal, the method introduces a computing unloading mechanism to reduce the delay and the overhead of the local terminal itself. The method is based on a D2D communication technology. In the scenario where mobile terminals are densely distributed, the local terminal can unload the computing tasks at the same time to several surrounding idle terminals for processing. According to the method, a total overhead objective function is established in consideration of the task delay, energy consumption, and unloading fee.

Abstract: A method for selecting a content forwarding node in a vehicle ad-hoc network on the basis of service type, comprising the following steps: calculating a bandwidth occupation proportion factor according to the type of requested content information and a bandwidth occupation situation of a neighboring vehicle which can perform forwarding in a one-hop transmission range; defining a virtual arrival time of a candidate forwarding vehicle to an initial request vehicle according to the distance between the neighboring vehicle and the initial request vehicle and a driving speed of the neighboring vehicle; calculating a forwarding node impact factor according to the bandwidth occupation proportion factor and the virtual arrival time; and selecting a forwarding node according to the forwarding node impact factor.

Abstract: A multi-timescale voltage regulation method based on source-grid-load-storage multi-terminal collaboration of a power distribution network is disclosed, which comprises: establishing, based on a Petri network, a multi-mode switching control model based on voltage security event trigger to realize effective control of a global voltage; establishing multi-objective optimization taking into account a source-storage-load regulation cost and a network transmission loss to realize collaborative and dynamic control of controllable resources of a source terminal, a load terminal and a storage terminal in each operating mode; and establishing a source-storage-load multi-terminal collaboration-based distributed voltage control model based on voltage security event trigger over a short timescale by taking into account the problems of voltage magnitude being out of limit and voltage leap, and solving online an optimal control sequence of the source terminal, the load terminal and the storage terminal in a receding horizo

Abstract: Disclosed is a joint recommendation and cache optimization method based on a collaboration of multiple base stations, each base station in a target area randomly caching contents stored to the cloud, when each base station in the target area receives a to-be-responded content request, constructing the recommendation optimization sub-model, the access optimization sub-model and the cache optimization sub-model, then constructing and solving the optimization problem to obtain the target base station, optimal cached contents and the recommendation list that correspond to the to-be-responded content request. Through the technical solution of the present application, the total time delay of the system is minimized, and by jointly optimizing the recommendation decision, access decision and cache decision, the delay can be significantly reduced with a lower cache capacity, and more efficient for users, good experience.

Abstract: Disclosed is a joint recommendation and cache optimization method based on a collaboration of multiple base stations, each base station in a target area randomly caching contents stored to the cloud, when each base station in the target area receives a to-be-responded content request, constructing the recommendation optimization sub-model, the access optimization sub-model and the cache optimization sub-model, then constructing and solving the optimization problem to obtain the target base station, optimal cached contents and the recommendation list that correspond to the to-be-responded content request. Through the technical solution of the present application, the total time delay of the system is minimized, and by jointly optimizing the recommendation decision, access decision and cache decision, the delay can be significantly reduced with a lower cache capacity, and more efficient for users, good experience.

Abstract: A networked control method for primary frequency regulation of a new energy power station based on a source-grid-load-storage networked cloud decision control system platform comprises: determining, according to historical operating data of a new energy power station, primary frequency regulation predictive values of power generation units of the new energy power station; determining, according to the primary frequency regulation predictive values, optimal control sequences of inverters of the power generation units at different times based on a pre-established inverter active power model, wherein the optimal control sequences comprise multiple control quantities of active power of the inverters; marking the optimal control sequences at the different times with time scales, sending the optimal control sequences to executing devices of the power generation units, receiving the optimal control sequences, and determining whether to store or not store the optimal control sequences; and determining the control qua

Abstract: A method of image reconstruction for a cross-modal communication system is disclosed. The method reconstructs a damaged, lost, or delayed image signal during transmission by using complete haptic signals received by a receiving end in the cross-modal communication system, and further constructs a cross-modal interaction network with reference to an attention mechanism, thus solving the limitation of the conventional generation model that it can only be trained on paired samples. An image reconstruction device for a cross-modal communication system is also disclosed. By fully utilizing semantic correlation between different-modality data and realizing cross-modal generation from haptic signals to image signals for unpaired data, the present invention overcomes the difficulty in acquiring haptic-image signal pairs in the practical cross-modal communication system, and significantly improves the quality and class accuracy of generated image signals.

Abstract: A precision compensation mechanism of a full-automatic bidirectional flanging machine comprises an upper die X-direction fine adjustment assembly, an upper die Y-direction fine adjustment assembly, a lower die X-direction fine adjustment assembly and a lower die Y-direction fine adjustment assembly, wherein the fine adjustment assemblies each comprise N adjusting threaded sleeves, N adjusting screws and a threaded sleeve rotation driving device; wherein the N adjusting threaded sleeves are threaded into an upper opening or a lower opening of a C-shaped flanging beam along the length direction of an upper die or a lower die; the outer wall surfaces of the N adjusting threaded sleeves are threaded into the upper opening or the lower opening of the C-shaped flanging beam to form N screw thread pairs I with the thread pitch of P1.

Abstract: The present disclosure discloses a multi-task joint computing unloading and resource allocation method based on D2D (Device-to-Device) communication. For solving the problems that computing resources of a local terminal cannot complete all computing tasks on time when there are multiple computing tasks with delay requirements on a local terminal, the method introduces a computing unloading mechanism to reduce the delay and the overhead of the local terminal itself. The method is based on a D2D communication technology. In the scenario where mobile terminals are densely distributed, the local terminal can unload the computing tasks at the same time to several surrounding idle terminals for processing. According to the method, a total overhead objective function is established in consideration of the task delay, energy consumption, and unloading fee.

Abstract: The present disclosure discloses a multiscale point cloud classification method. The method includes the following steps: acquiring 3D unordered point cloud data; performing feature extraction and classification on the acquired point cloud data using a pre-trained parallel classification network to obtain an output result, wherein the parallel classification network includes a plurality of basic networks with the same structures; and fusing the output results of the parallel network using a pre-trained deep Q network to obtain a final result of point cloud classification. The present disclosure can improve the accuracy and robustness of point cloud classification.

Abstract: The present disclosure discloses a high-precision point cloud completion method based on deep learning and a device thereof, which comprises the following steps: introducing dynamic kernel convolution PAConv into a feature extraction module, learning a weight coefficient according to the positional relationship between each point and its neighboring points, and adaptively constructing the convolution kernel in combination with the weight matrix. A spatial attention mechanism is added to a feature fusion module, which facilitates a decoder to better learn the relationship among various features, and thus better represent the feature information. A discriminator module comprises global and local attention discriminator modules, which use multi-layer full connection to classify and determine whether the generated results conform to the real point cloud distribution globally and locally, respectively, so as to optimize the generated results.

Abstract: The present disclosure discloses an image haze removal method and apparatus, and a device. The method includes: acquiring a hazy image to be processed; and obtaining a haze-free image corresponding to the hazy image by inputting the hazy image into a pre-trained haze removal model. The present disclosure uses the residual dual attention fusion modules as basic modules of the neural network, so that each feature map can obtain pixel features while enhancing the global dependence, thus improving the image dehazing effect.