89results about How to "Easy to train" patented technology

This utility model relates to the field of apple seedling propagation technology and discloses an apple seedling propagation and cultivation device, including a cultivation box. A cultivation mechanism is provided inside the cultivation box, and a moisture-retaining mechanism is provided at the bottom of the cultivation box. The cultivation mechanism includes a cultivation component, a sealing component, a locking component, and a pulling component. The cultivation component is disposed inside the cultivation box. With this apple seedling propagation and cultivation device, through the cultivation mechanism, workers can use locking blocks to engage the cultivation troughs with the top of the frame, increasing the stability of the cultivation troughs after installation within the cultivation box. Simultaneously, the equidistantly spaced cultivation troughs make the space between multiple cultivation troughs more rational, which is more conducive to the cultivation of apple seedlings. Furthermore, the sliding plate, driven by a compression spring, moves the locking rod backward, engaging it with the cultivation box and the closed doors on the left and right sides, increasing the stability of the closed doors after closure.

The utility model relates to the technical field of morchella esculenta production, and discloses a morchella esculenta culture shelf which comprises two supporting plates which are distributed up and down at an interval and a side shelf with the two ends fixedly connected with the supporting plates, culture assemblies for placing morchella esculenta stock seeds are arranged between the supporting plates, and the two culture assemblies are arranged in a front-and-back symmetry mode. The two culture assemblies are supported and fixed through the supporting assemblies and the supporting plates. The stock culture shelf has the advantages that a one-groove culture mode is adopted, the placing space of the culture shelf is reasonably utilized, stock seeds cannot be stacked when placed and cultured, taking out or placing operation of each stock seed becomes simple and fast, meanwhile, the air circulation on the shelf is guaranteed, and meanwhile the culture shelf is convenient to use. And the hypha development condition of the stock seeds can also be visually observed, so that the stock seeds are convenient to culture, load and take, and the orderliness and efficiency of production are improved.

The automobile front door cover size detection device of the present disclosure comprises a part storage table for placing parts, a part grabbing hand for grabbing parts, a support component for fixing and clamping parts, a detection system comprising a data acquisition element, a data calculation system and a data display system, the data acquisition element is used for acquiring part data, the data calculation system is used for determining part size change value according to part data, and the data display system is used for outputting part size change value; a collaborative robot is used for moving the part grabbing hand, and the collaborative robot is used for moving the data acquisition element; a control system is in communication connection with the part grabbing hand, the detection system and the collaborative robot, the control system is used for outputting detection instructions, the detection instructions comprise gripping instructions, part moving instructions and scanning instructions, the part moving instructions are used for controlling the collaborative robot to move the parts through the part grabbing hand, and the scanning instructions are used for controlling the collaborative robot to move the data acquisition element to scan the parts to acquire part data.

The application discloses a preparation method of carbon dot material based on blue bacteria, and the carbon dot material is prepared by taking blue bacteria photosystem pigment as raw material. The semiconductor fluorescent carbon dot prepared by the method has stable properties, has wide spectrum absorption to sunlight, can convert ultraviolet light, visible light and infrared light into red light, and has good photoelectric performance. The application further discloses a preparation method of a blue bacteria-based semiconductor fluorescent carbon dot-blue bacteria hybrid, and application of the carbon dot material based on blue bacteria in crop production.

This invention discloses a baseline removal method for multi-channel magnetocardiogram (MCC) signals based on wavelet low-frequency reconstruction and morphological secondary smoothing. The invention includes: acquiring multi-channel MCC signal data; performing power frequency notch filtering on each channel to suppress the power frequency fundamental frequency and at least one harmonic component; performing configurable IIR low-pass filtering on the notched signal; setting a discrete wavelet transform extension mode and performing discrete wavelet decomposition on the filtered signal; selectively retaining decomposition coefficients based on the retention layer set and setting the remaining coefficients to zero, then performing inverse wavelet reconstruction to obtain an initial baseline estimate; performing morphological secondary smoothing on the median-smoothed baseline signal to obtain a corrected baseline signal; subtracting the corrected baseline signal from the filtered signal to obtain the baseline-corrected channel signals; and assembling and outputting a multi-channel baseline-corrected signal matrix. This invention can improve the stability and cross-channel consistency of baseline estimation for multi-channel MCC signals, reduce the risk of over-correction and under-correction, and is applicable to multi-channel batch processing scenarios.

The utility model discloses a kind of citric acid fermentation fine filter devices, belong to citric acid production technical field.Device is composed of fermentation unit and fine filter unit: fermentation unit contains seed tank and fermentation tank, the feed inlet of seed tank is connected with syrup turbid liquid pipe, the feed inlet of fermentation tank is connected with syrup clear liquid pipe and syrup turbid liquid pipe, seed tank and fermentation tank are configured with steam, compressed air, defoaming agent, emptying and circulating water pipeline.Fine filter unit first removes most solid impurities with primary plate-and-frame filtration, then ceramic membrane fine filtration and anion and cation exchange impurity removal.Two-stage filter-pressing acid residue is dried after being thickened and then filter-pressing to be feed, and two-stage filtrate respectively enters subsequent process.The device can greatly reduce labor cost and failure rate, improve citric acid purity and yield, and meet the needs of large-scale, high-quality and green production.

This invention provides an intelligent evaluation method for rockburst hazard at working faces based on principal component analysis-probabilistic neural networks, belonging to the field of rockburst hazard evaluation technology. An evaluation index system is established and data is collected by combining rockburst hazard influencing factors, drill cuttings monitoring, critical stress index monitoring, and actual field conditions. Principal component analysis (PCA) is used to simplify the original evaluation index data, resulting in comprehensive evaluation index data containing information from the original evaluation indicators. The comprehensive evaluation index data is divided into a training set and a test set. A probabilistic neural network (PNN) is used to train the evaluation model on the data in the training set, and then the performance of the evaluation model is tested using data from the test set, and its accuracy is calculated. If the accuracy is greater than or equal to 90%, the evaluation model is considered acceptable; if the accuracy is less than 90%, it is considered unacceptable, and the PNN's smoothing factor needs to be modified and retrained until the accuracy of the evaluation result meets the set value.

The application discloses a method for repairing heavy metal contaminated soil by combining plants and microorganisms, and comprises the following steps: planting alfalfa seeds in heavy metal contaminated soil; inoculating lead and zinc copper resistant bacteria L1 into the heavy metal contaminated soil after 30 days when the seeds germinate; and harvesting the alfalfa after the growth period of the alfalfa ends. The application has a very obvious effect on adsorbing lead, zinc and copper heavy metals in the soil. The plant raw material is low in cost, the preparation process is simple and easy to operate, and the bacterial liquid is easy to culture. The application has small investment for removing heavy metals and no secondary pollution.

The application relates to a video instance segmentation method based on a dynamic convolution decomposition of a lightweight attention mechanism; the method first inputs a video frame, and a backbone network independently extracts a feature map of each frame in the video; then the feature map extracted by the backbone network enters an HQT encoding and decoding module, the position change of an instance in each frame is accurately positioned by combining an output head, and three prediction branches are used to supervise model training; in the application, a convolution layer of an encoding network adopts dynamic convolution decomposition, a more compact model is obtained, model training is easier, the required parameter quantity is greatly reduced, the training speed is improved, and the training time is shortened; a lightweight HQT encoding and decoding module is provided, which helps to reduce model parameters and improve efficiency; the loss function is improved, model training is more stable, the convergence speed and convergence precision are improved, and the problems of imbalance between foreground and background in samples and imbalance between foreground categories under a long tail condition are relieved.

The invention relates to the technical field of online car-hailing supervision, in particular to an online car-hailing supervision method and system. Comprising the steps of collecting data and performing data preprocessing to obtain an initial data set; inputting the initial data set into an urban online car-hailing saturation judgment model; outputting a judgment result through a data input layer, a data prediction layer, a data training layer and a judgment output layer in sequence; the judgment result is used for representing whether the online car-hailing number of the city is saturated or not; the data prediction layer predicts the initial data set through an RVM prediction model and an LSTM prediction model to obtain a prediction data set; the data training layer comprises a trained BP neural network; and the BP neural network carries out training on the prediction data set and outputs a training result. The invention provides an urban online car-hailing saturation judgment model which can perform prediction according to historical data of a city and judge whether the number of urban online car-hailing cars is saturated or not.

This utility model discloses a tooling for the impregnation process of a square exciter stator, including a hoisting tool and an impregnation fixture. The hoisting tool includes a main lifting beam, two eye bolts, and several stops. The main lifting beam includes a rectangular tube and a pair of support plates symmetrically arranged on both sides of the top of the rectangular tube. A sealing plate is fixed to each end of the main lifting beam. The inner surfaces of the pair of support plates are machined into bevels and fixed to the top sides of the rectangular tube. The two eye bolts are fixed to the outer surfaces of the sealing plates at both ends of the main lifting beam. Several stops are fixed to the top surface of the rectangular tube by bolts with a spacing adapted to the thickness of the exciter stator. The impregnation fixture includes an outer frame, two stator support beams fixed to the bottom of the outer frame, a lead wire fixing bracket fixed to the top of the outer frame, and four hooks. This utility model not only solves the problems of laborious and risky hoisting, but also the problems of long dripping time and easy paint accumulation during impregnation.

The application discloses a license plate quality judgment method based on semi-supervised learning, comprising: obtaining a license plate image to be judged, and obtaining a license plate character local image according to an existing license plate detection algorithm; inputting the license plate character local image into a pre-trained deep neural network model, wherein the deep neural network model comprises a feature extraction module and a license plate quality prediction module; using the feature extraction module to perform feature extraction on the license plate character local image to be judged, and obtaining a high-level feature map comprising the license plate character local image; using the license plate quality prediction module to process the high-level feature map, and obtaining a license plate quality index of the license plate character local image; the network model of the license plate quality judgment method has strong universality, is faster in recognition, has stronger robustness for low-quality license plate images with dirt, adhesion, character loss and inaccurate positioning, does not need additional labeling of license plate quality, and is more convenient in model training.

The invention provides a remote semi-physical simulation error compensation method and system. The remote semi-physical simulation error compensation method comprises the following steps: S1, acquiring air situation information of our radar guidance product and an attacking target; s2, predicting the maneuvering type of the strike target according to the our radar guidance product and the air situation information of the strike target; s3, according to the maneuvering type of the strike target, track prediction is carried out on the strike target; and S4, according to the maneuvering type and a trajectory prediction result, finishing compensation of a remote simulation error. According to the method, the technical effect of reducing simulation errors caused by remote data interaction time delay in radar guidance product cluster co-simulation is achieved.

The present application relates to the technical field of rehabilitation equipment, and particularly relates to a whole-body joint rehabilitation training device and a control method thereof. The method of the present application comprises: collecting joint angles, IMU data and electromyographic signals when a user performs a plurality of standard actions; determining a joint range of motion risk level of a target joint according to the joint angles; determining a compensatory behavior risk level and a synergy efficiency score of the target joint according to the joint angles, the IMU data and the electromyographic signals; determining a training mode according to the compensatory behavior risk level, wherein the training mode at least comprises a general mode and a mode for compensatory behavior; determining a training parameter according to the joint range of motion risk level, wherein the training parameter comprises resistance; adjusting the training parameter according to the synergy efficiency score; and outputting the training mode and the adjusted training parameter. The method of the present application can improve rehabilitation efficiency and reduce compensatory behavior during training.

The utility model discloses a cell culture plate, which comprises a bottom box plate, the bottom box plate comprises a box body, a culture groove, a hook plate and a cover plate, the culture groove is arranged in the box body, the inner wall of the box body is connected with the hook plate, the cover plate is arranged in the culture groove, and the top of the cover plate is provided with a handle groove; the base plate comprises a frame body, a keel, a sleeve and a clamping plate, the sleeve is arranged in the frame body, the sleeve is connected with the frame body through the keel, and the lower portion of the frame body is connected with the clamping plate; and the box plate is arranged above the bottom box plate. According to the cell culture plate, through the arrangement of the cover plate, the culture grooves are conveniently blocked, so that culture of cells in the single culture groove is facilitated, mutual interference among the cells is avoided, the height of the culture grooves can be increased through the base plate, subsequent operation of cell culture is facilitated, and culture and development of the cells are facilitated.

The application discloses an aircraft cluster cooperative non-cooperative target capturing method driven by entropy regularization, applies the entropy regularization thought to reinforcement learning, optimizes the updating method of the parameters of an evaluation network and a policy network, so that a UAV can obtain a more optimal maneuvering strategy, and the UAV performs actions according to the maneuvering strategy, and realizes the interception of an attacking aircraft. The aircraft cluster cooperative non-cooperative target capturing method driven by entropy regularization can not rely on a deterministic strategy in the interception process of multiple UAVs, but adopts a random strategy, avoids the training from falling into a local optimal point, and the actions of the aircraft are randomized as much as possible while the aircraft completes the task, so that the exploratory and robustness can be practically improved.

The utility model discloses a mechanical swing arm structure for leg rehabilitation, which mainly comprises a push rod support, a small swing arm, a mounting rack and the like, when an electric push rod is retracted, the push rod support is perpendicular to the small swing arm, when the electric push rod is ejected, the small swing arm can be freely adjusted within the range of 0-25 degrees, a bicycle type pedal mechanism is arranged at the bottom of the mounting rack, and the surface of a pedal is antiskid. A shank supporting part is arranged on the rear portion, and a shank hanging mechanism, a motor and a displayer are arranged on the installation frame. A user carries out training through a pedal mechanism, a motor provides power or damping for the device, a displayer feeds back training data, motor parameters can be set, an electric push rod is electrically controlled, accurate regulation and control are achieved through a button or a remote controller, and compared with traditional linear adjusting equipment, the structure is simpler and more convenient to operate; the damage risk of external force to equipment is effectively reduced, the structural stability is improved, and the actual requirements of leg rehabilitation training can be better met.

The invention discloses a large model training and reasoning-oriented containerized operation method and computing equipment, and the method comprises the steps: receiving a training starting request or reasoning starting request for a large model, and determining a training parameter corresponding to the training starting request or a reasoning parameter corresponding to the reasoning starting request; for the training starting request, creating a trainer instance and introducing training parameters, starting a training container through the trainer instance based on the training parameters, and running a training program through the training container so as to execute a large model training task based on the training program; and for the inference starting request, creating an inference device instance and transmitting inference parameters, starting an inference container based on the inference parameters through the inference device instance, and running an inference program through the inference container so as to execute a large model inference task based on the inference program. On the basis, training and reasoning tasks of the large model can be efficiently and conveniently executed in a containerized environment, and an embeddable containerized operation capability can be provided for a model development platform, so that other platforms can conveniently and rapidly integrate the development capability of the large model.

The invention discloses an in-service nurse nursing behavior intelligent analysis and training system based on artificial intelligence, and relates to the technical field of nursing behavior analysis. The information acquisition module is used for acquiring past nursing behavior information and acquiring real-time nursing behavior information of an in-service nurse; according to the invention, the reference library is established based on the previous nursing behavior information through the set reference establishment method as a reference, the nursing name corresponding to the real-time nursing behavior information is judged through the behavior matching method, and whether the nursing behavior of the nurse is standard is judged through the behavior analysis method in combination with the nursing name. And when the behavior is not standard, to-be-optimized information is generated through a behavior optimization method and sent to nurses, intelligent analysis behaviors are integrally provided, training is provided when the behaviors have defects, the compliance and skill gap in actual operation is continuously and objectively evaluated through systematic analysis and training, near-real-time error correction and decision support is provided, and the accuracy and the reliability of the system are improved. And the safety and the efficiency are instantly improved in key nursing links.

This invention relates to a flood disaster assessment method based on the fusion of remote sensing imagery and GIS data, belonging to the field of flood disaster monitoring and assessment technology. This method solves the problems of insufficient timeliness, limited coverage, and the inability of a single data source to comprehensively reflect the disaster situation in traditional assessment methods. Its technical solution mainly includes utilizing GIS data to provide spatial semantic background support, employing an improved UMNet++ model to accurately extract water body information before and after the flood from Sentinel-2 remote sensing imagery, and achieving quantitative assessment and hierarchical visualization of disaster severity through multi-source data fusion and spatial overlay analysis. The beneficial effects of this invention are that it significantly improves the accuracy and comprehensiveness of flood disaster assessment, enabling rapid and scientific disaster response and decision support.

The application discloses a style transfer image quality evaluation method, which comprises the following steps: inputting content and result images into a content preservation quality prediction network for training, outputting a content preservation quality score and a content preservation weight vector, and obtaining a content preservation quality prediction network training model; inputting style and result images into a style similarity quality prediction network for training, outputting a style similarity quality score and a style similarity weight vector, and obtaining a style similarity quality prediction network training model; inputting the content preservation quality score and the weight vector and the style similarity quality score and the weight vector into an overall visual quality prediction network for training at the same time, outputting an overall visual quality score, and obtaining an overall visual quality prediction network training model; and obtaining the content preservation quality score, the style similarity quality score and the overall visual quality score of the result image by using the three training models during testing.

The invention belongs to the field of protein fluorescence intensity prediction, and particularly relates to a protein sequence fluorescence intensity prediction method, system and device and a storage medium, and the method comprises the steps: carrying out the feature extraction of a training data set, combining random noise, training a generator and a discriminator, obtaining a generative adversarial model, supplementing data through the generative adversarial model, and obtaining a protein sequence fluorescence intensity prediction result. The method comprises the steps of obtaining a prediction data set, then extracting semantic features of the prediction data set, performing local mode enhancement and fusion on hidden layer features to obtain final features, inputting the final features into a prediction network to train the prediction network to obtain an optimal prediction network, and inputting an actual protein sequence into the optimal prediction network. The fluorescence intensity corresponding to the actual protein sequence is obtained. The method has the effect of improving the accuracy of predicting the fluorescence intensity of the protein sequence.

This application discloses a method, apparatus, device, and storage medium for enterprise data processing based on a large model. The method, relating to the field of data processing technology, includes: determining multiple sets of enterprise data corresponding to the enterprise to be predicted in an enterprise information database, wherein each set of enterprise data corresponds to a different data type, and the enterprise information database is constructed by acquiring original enterprise data from different enterprises; for each set of enterprise data, inputting the original data of the target enterprise in the set of enterprise data into an expert model corresponding to the enterprise data set in a pre-trained large model to obtain feature vectors; determining the vector weights corresponding to each feature vector based on the gate network in the large model, and determining and outputting the enterprise data prediction result based on each feature vector and its weights. This application improves the accuracy of the model in predicting enterprise loan intentions.

The application relates to a text sensitive word library extraction method, device and equipment based on a neural network model. The method comprises the following steps: constructing a sensitive word library extraction model; the sensitive word library extraction model comprises a self-defined rule algorithm, a character-by-character segmentation algorithm, a Chinese word segmentation algorithm and an N-gram algorithm; the sensitive word library extraction model is pre-trained according to a training data set; the pre-trained sensitive word library extraction model is used for word segmentation extraction and word library quantitative analysis on the text to be extracted; the full-amount word library is subjected to field structure design; the pre-trained sensitive word library extraction model is iteratively trained according to the designed word library; after the sensitive word extraction is carried out by using the trained sensitive word library extraction model, data analysis is carried out according to a word segmentation frequency filtering rule, a word segmentation character quantity filtering rule and a word segmentation type filtering rule, and a sensitive word library is obtained. The method can improve the sensitive word extraction accuracy and efficiency.

The invention relates to an intelligent system for aortic balloon occlusion training, which comprises a human-shaped shell, an artery simulation structure and a blood simulation mechanism, the artery simulation structure is arranged in the human-shaped shell and comprises an artery channel used for simulating human artery blood vessels, and the tube wall of the artery channel is provided with a detachable puncture part; the puncture part is made of an elastic recoverable material, the puncture part is used for a puncture needle to penetrate through, and the puncture part can be tightly attached to the puncture needle when the puncture needle is inserted; the human-shaped shell is provided with a puncture needle entering area corresponding to the puncture part, the puncture needle entering area is provided with a simulated skin layer made of an elastic recoverable material, and a puncture needle penetrating through the puncture needle entering area can be inserted into the corresponding puncture part; the blood simulation mechanism is connected with the end, close to the neck, of the artery channel and used for providing blood simulation liquid for the artery channel.

The utility model belongs to the technical field of preschool education, and particularly relates to a preschool education picture watching and object identifying device, which solves the problems of incapability of efficiently guiding and inconvenience in carrying in the prior art, and comprises a picture board, a circuit module, a broadcasting module and a power supply module are respectively and fixedly arranged on the picture board, and a plurality of buttons are arranged on the circuit module. The circuit module is electrically connected with the broadcasting module, and the circuit module and the broadcasting module are both electrically connected with the power supply module; the shell is of a cavity structure, a sliding block is arranged on the mounting plate in a sliding mode, a connecting plate is fixed to one side of the sliding block through a screw, through the arrangement of the mounting plate, the sliding groove and other structures, an infant can autonomously operate a rotary knob to adjust the height after mounting, and therefore the infant can autonomously use the device, and the practicability of the device is improved. And moreover, operable parts are added, so that the device can train and guide the infant more easily.

This invention provides a compound microbial agent, its preparation method, and its application in the control of plant root rot, belonging to the field of plant disease biological control technology. The compound microbial agent prepared by combining *Bacillus mohaiweiii* DS-L23, *Pseudomonas fluorescens* DS-PZ-L7, and *Pseudomonas aeruginosa* DS-L22 exhibits significant inhibitory effects against various plant pathogenic fungi of *Codonopsis pilosula*, including *Fusarium oxysporum*, *Fusarium solani*, *Fusarium trifidum*, *Pseudomonas heterostemum*, and *Alternaria alternata*. This compound microbial agent has advantages such as good stability and ease of cultivation. Furthermore, the construction method of the compound microbial agent is simple and easy for industrial mass production. After application, the compound microbial agent provided by this invention has a good control effect on *Codonopsis pilosula* root rot and can promote the growth of *Codonopsis pilosula* plants, reducing the use of chemical pesticides while improving the control effect.

The invention discloses a speech emotion recognition method, system and device based on MFCC and LFCC feature fusion and a storage medium, and belongs to the technical field of speech emotion recognition. The method comprises the following steps: preprocessing an input voice signal, and respectively extracting Mel frequency cepstrum coefficient characteristics and linear frequency cepstrum coefficient characteristics; inputting the extracted MFCC features and LFCC features into a long short-term memory network for time sequence feature fusion; inputting the fusion feature vector into an improved ECAPA-TDNN model for sentiment classification, wherein the improved model uses an efficient channel attention module to replace a compression excitation module in an original model; and finally outputting an emotion recognition result. According to the method, the MFCC and LFCC features are fused, the defect that the resolution of a single MFCC feature in a high-frequency area is insufficient is overcome, the LSTM network and the improved ECAPA-TDNN model are combined, deep fusion and efficient modeling of the voice time sequence and the frequency spectrum features are achieved, and the model complexity is effectively reduced while the high recognition accuracy is guaranteed.

The application discloses a color image quantization step estimation method based on frequency clustering prior knowledge, first, the color image obtained in advance is pretreated, and the image is converted from a spatial domain to a frequency domain; second, an improved Res2Net-C network structure is constructed to obtain quantization step information in the frequency domain; finally, a multi-channel convolution is introduced to assist in estimating the quantization step of the chroma channel. The application converts the image from the spatial domain to the frequency domain through the pretreatment operation, facilitates the network to explore the traces of the quantization step in the JPEG image, and significantly improves the accuracy of estimating the quantization step; the new Res2Net-C network structure can explore the multi-scale information of the image, thereby improving the accuracy of estimating the quantization step; compared with the traditional method for estimating the quantization step, the application has high accuracy and is easier to train.

The utility model provides a kind of sit and pedal trainer, including support column, rocker and support component, rocker can swing and / or swing left and right relative to support column, the upper end of rocker is provided with holding part and / or the lower end of rocker is provided with tread part;Support component includes support mounting, and first support piece and second support piece;First support piece includes first support arm, and first support foot, and second support arm is rotationally arranged on support mounting, and second support foot, second support piece includes second support arm, and second support foot, and first support foot and second support foot can be mutually close away in the rotation process of first support arm and second support arm.Support mounting is set on support column, and first support arm and second support arm rotation process can make first support foot and second support foot mutually close away.The utility model can be stably reliable to make support column stand in unfolded state, in folding state, the floor area required when sit and pedal trainer is stored can be reduced, to facilitate storage.