60results about How to "Solve learning problems" patented technology

The invention relates to an inclined shaft (roadway) transportation human-oriented safety monitoring and training system in a mining area and a manufacturing method; the inclined shaft (roadway) transportation human-oriented safety monitoring and training system comprises an inclined shaft (roadway) transporting environment, power-driven car arrestors, a safety gate, a voice communication device, a stay wire interlocking device, an infrared human body detecting sensor, a video monitoring system, a PLC (Programmable Logic Controller) and a configuration software monitoring system; the power-driven car arrestors, the safety gate, the voice communication device, the stay wire interlocking device, a camera and the sensor are arranged in an inclined shaft (roadway); all the devices are connected with the PLC by using a field bus; the PLC is used for carrying data processing and logical operation; an optical fiber is used for transmitting data to a computer; the configuration software operates to carry out dynamic display and monitoring; and therefore, the closed inclined shaft transportation is realized, and vehicle passing without permission of person passing and person passing without permission of vehicle passing are ensured. The mechanical parts are designed according to a reducing scale, and electrical parts are actually industrial components, thereby the inclined shaft (roadway) transportation human-oriented safety monitoring and training system is really represented; and the inclined shaft (roadway) transportation human-oriented safety monitoring and training system is reasonable in design, compact in structure, advanced in technology and convenient to use and is suitable for being used for experiment training in colleges and universities and training for workers and staffs of enterprises.

The invention provides a BP neural network-based multimodal motion method and system of robot fish. The method comprises the steps of building a CPG model; carrying out dynamic modeling on robot fish with pectoral fins at four joints, inhibiting pectoral fin CPGs by tail fin CPGs in a one-way manner, determining left and right input stimuluses, downlink and uplink phase coupling coefficients, uplink and downlink coupling coefficient weights and CPG frequencies corresponding to various joints by using nonlinear oscillator models as CPG neurons; building a BP neural network model; obtaining variations of joint angles on the basis of the CPG model, storing variation values of the joint angles as data packets for carrying out BP neural network training, and transmitting the trained data to a controller of the biomimetic robot fish; and driving swinging of various joints by using CPG signals and carrying out swimming and turning motions of the robot fish. According to the BP neural network-based multimodal motion method and system of the robot fish, multimodal motions of the robot fish are learned by using the BP neural network, so that the target of learning the multimodal motion processes of the robot fish through the BP neural network is finally achieved and the autonomy and the adaptability of a robot fish system are improved.

The invention relates to a handwriting recognition child intelligent learning system based on an intelligent terminal. The handwriting recognition child intelligent learning system comprises the intelligent terminal with a touch screen, a feature extraction module, a character knowledge base, a character recognition module, a learning mode selection module and a voice output module, wherein the feature extraction module, the character knowledge base, the character recognition module, the learning mode selection module and the voice output module are arranged in the intelligent terminal; the feature extraction module is connected with the touch screen and the character recognition module; the character knowledge base is connected with the character recognition module and the learning mode selection module; the voice output module is connected with the character recognition module and the learning mode selection module. Compared with the prior art, the handwriting recognition child intelligent learning system has the advantages that by the utilization of the handwriting input function of the intelligent terminal such as a tablet personal computer and an intelligent phone, the problem that patients have no time or do not have enough time to coach children in learning is solved, the handwriting recognition child intelligent learning system is not limited by learning time and places, and economic burdens are greatly reduced.

The invention discloses a blast furnace smelting expert system built based on a pattern recognition technology, which comprises a server, an operator workstation, six sets of PLCs (programmable logic controllers) (respectively used for material preparation below grooves, material distribution at the top of a blast furnace, water granulated slag control, dust removal control, coal powder injection control and water system control), a distributed control system, an IPC (industrial personal computer) for communication, a state classification module, a state diagnosis module, a state evaluation module, a state adjusting module and a state display module, wherein the server is provided with a database, the state classification module, the state diagnosis module, the state evaluation module and the state adjusting module are arranged in the server with the database, and the state display module is arranged in the operator workstation. Meanwhile, the invention also discloses a method for building a blast furnace smelting expert system based on a pattern recognition technology, which comprises the following steps of data access, state classification, state diagnosis, state evaluation, state adjustment, state display and execution. The system and method disclosed by the invention have the characteristics of standard and objective evaluation, good integrated performance, rich knowledge base and real-time updating, and can be widely applied to the technical field of blast furnace process control.

The invention aims at providing a method and device for sending and receiving a message. The method for sending and receiving the message comprises firstly, receiving an internet protocol message to be sent from a source virtual machine, then removing a target media access control address in the internet protocol message in order to generate a virtual extensible LAN message, and finally sending the virtual extensible LAN message to a switch corresponding to a target virtual machine according to a preset address list recorded with the mapping relation between the internet protocol address of the target virtual machine and the internet protocol address of a host machine. The media access control address learning problem is solved, at the same time, the virtual extensible LAN message gettingrid of the redundant media access control address is employed, the message is forwarded completely based on the internet protocol address of the target virtual machine, the bandwidth utilization rateis effectively increased, and good scalability is achieved at the same time.

The invention discloses a constructing method for a decoupling controller of a bearingless permanent magnet synchronous motor with five degrees of freedom. Three Clark inverse transformers are respectively connected in front of three corresponding current tracking inverters in series, and the three current tracking inverters and one linear power amplifier are respectively connected in front of the synchronous motor and a load model thereof in series to form a complex controlled object; and a support vector machine inverter with 6 input nodes and 7 output nodes is constructed by a support vector machine with 17 input nodes and 7 output nodes and 11 integrators, a pseudo-linear system is formed, a corresponding pseudo-linear subsystem controller is designed, a linear closed-loop controller is constructed, and the decoupling controller of the synchronous motor is constructed by the linear closed-loop controller, the support vector machine inverter, the three Clark inverse transformers, the three current tracking inverters and one linear power amplifier. Nonlinear dynamic decoupling control on the rotor displacement and the rotating speed of the bearingless permanent magnet synchronous motor with five degrees of freedom can be realized.

The invention discloses a model detection method based on feature selection, and the method comprises the steps: cutting an original data set; training a model through employing a training set; predicting the training set and a verification set through employing the model, and obtaining a prediction error; deleting features, obtaining a new training set, predicting the new training set through the model, and obtaining a prediction error; assigning j to (j+1), returning to the last step, and executing the next step till the value is C; calculating prediction error distances between the new training set and the training set; ordering the distances, and searching features corresponding to G minimum distances; storing sequence numbers of the features into a feature deleting feature sequence, and deleting features in the training set and a verification set; assigning C to (C-G), returning to a second step, and executing the next step till C is not greater than G; obtaining a sequence number K according to the prediction error, and deleting former (K-1) features from the training set and a testing set; training the new model through employing the training set with the features being deleted, predicting the testing set with the features being deleted through employing the new model, and obtaining the prediction error.

The invention discloses a learning problem answering method and system. The method comprises the following steps: obtaining problem information uploaded by students; sending answering invitations to students; after answering invitation responses of the students are received, obtaining answering information of the students; and displaying correct answering information to the students. The learning problem answering method and system solves the problem of incapability of detailed answering of learning problems which the students encounter after class and the homework plagiarizing problem of the students and enables the students to answer the problems for each other so as to realize common progress.

The invention discloses a control method of MAC address study and forwarding equipment. The method comprises the steps of receiving messages, and according to the attribute information of the messages, judging whether to conduct MAC address study according to the messages , wherein the attribute information comprises one of unicast, multicast and broadcast. By the adoption of the control method of the MAC address study and the forwarding equipment, the problems that when a fixed mode is used for controlling MAC address study, the MAC address study is single in mode, and the MAC address study is restricted are solved. The control method of the MAC address study can be controlled according to different requirements of the attribute information, can avoid impact of a large amount of broadcast or multicast messages existing in a network on system MAC address capacity, enables a system to study useful MAC address on the condition that capacity is limited, and improves system performance.

The embodiment of the invention discloses a cross-media multi-view imperfect label learning method. The method comprises the following steps: preprocessing an imperfect label training set, and extracting two groups of features of the training set from two view angles which are conditionally independent; training the two groups of the features and the original imperfect labels in the training set in a new multi-label two-view flexible support vector machine, so as to obtain a group of training parameters; reclassifying the training set which is independent in the two view angles according to the two groups of the training parameters, so as to obtain two groups of independent classification results of the training set in the two view angles; randomizing the two groups of the classification results to obtain two groups of independent randomized classification results of the training set in the two view angles; acquiring new imperfect labels for the training set through a group of new supplementing and denoising algorithms; and terminating the iterative process until the change between the obtained new imperfect labels for the training set and the original imperfect labels is smaller than the set threshold value.

The invention discloses a keyboard-based operation processing method, device and equipment and a medium, and relates to the technical field of human-computer interaction. The keyboard-based operationprocessing method comprises the following steps of receiving a key operation which is used for triggering the generation of the key information; determining the corresponding key coordinate data according to the key information; determining a geometric figure according to the key coordinate data, and identifying the geometric figure as a keyboard gesture corresponding to the key operation; and determining the trigger action of the preset function as a target action based on the mapping relation between the keyboard gesture and the preset function, and executing the target action to display anoperation result corresponding to the key operation. According to the method, the corresponding target action is executed by utilizing the recognized keyboard gesture, so that a user can trigger theequipment or software in the equipment to execute the corresponding action through the simple keyboard gesture, the problem of high learning and memory shortcut key cost of the user in the prior art is solved, and the operation efficiency of the user is improved.

The present invention discloses a robot quasi-periodic motion demonstration learning method. The method employs classification, decomposition, modeling and synthetic technologies in order to achieve learning and generalization reproduction functions of robot quasi-periodic motion, namely, a classification method is employed to determine that a motion track is periodic motion, non-periodic motion or quasi-periodic motion; a decomposition algorithm is employed to decompose the quasi-periodic motion to the periodic motion and the non-periodic motion; the periodic motion and the non-periodic motion are subjected to modeling learning and prediction, and each component after learning is combined to a new quasi-periodic motion according to an equation of definition of the quasi-periodic motion. The robot quasi-periodic motion demonstration learning method successfully represents the complex motion track-quasi-periodic motion of the robot in a non-structural environment to effectively solve the technical problems that the class of flexible track traditional automatic planning system and method is single in function and bad in universality.

The invention discloses a blockchain architecture based on machine learning competition. Individuals seeking machine learning solutions are allowed to issue machine learning tasks, nodes in the blockchain architecture can select tasks and carry out machine learning locally, and other nodes are randomly selected as model verifiers at a certain probability and are responsible for ranking machine learning models. After respective ranking is given, verifiers reach consensus of final ranking under a consistency protocol and submit new blocks to the block chain immediately, and the new blocks include transaction information, machine learning tasks, winning models and the like. According to the invention, a blockchain architecture is established based on machine learning competition; computing power resources wasted on the hash problem in the past are put into machine learning with practical significance, and effective utilization of the resources is realized. In the process of completing themachine learning task, a machine learning model and a data set storage library are also created, and continuous optimization of the machine learning model and the algorithm is promoted.

The invention discloses an incremental software defect prediction method, system and device and a storage medium, and the method comprises the steps: obtaining a first training set, carrying out the preprocessing of the first training set, and obtaining a second training set; based on the second training set, constructing a cost-sensitive LightGBM model; a cost-sensitive LightGBM model is used as a base classifier to construct an incremental prediction model, and the total training sample number and the total defect sample number are calculated; calculating a classification threshold according to the total training sample number and the total defect sample number; predicting a to-be-predicted sample by using the incremental prediction model to obtain a final prediction value; and judging the category of the to-be-predicted sample based on the classification threshold and the final prediction value. According to the method, a new base classifier can be added on the basis of an original model, historical samples do not need to be repeatedly trained while new data are learned, the model calculation amount can be reduced, the storage space is saved, the training efficiency is improved, the model is continuously improved, and the accuracy of software defect prediction is improved.

The invention provides a method for obtaining an artistic flower arrangement scheme. The method comprises the following steps of establishing a material library, wherein the material library comprisestemplates of flower materials, flower arrangement containers and decorative materials, stored in layers; according to hand painted design sample drafts, selecting the corresponding flower materials,containers and decorative materials from the material library to be put into the corresponding layers of electronic design initial drafts, and adjusting the contents of the layers; and combining the electronic design initial drafts to obtain the final artistic flower arrangement scheme. According to the technical scheme provided by the method, the electronic flower arrangement material library istaken as a basis and a computer application technology is taken as a guarantee, so that during flower arrangement practice and flower arrangement teaching, the flower materials, the containers and thedecorative materials can be selected from the flower arrangement material library directly through a computer, and the flower materials, the containers and the materials are reasonably combined according to a certain artistic requirement to obtain the satisfied flower arrangement scheme. Due to adoption of electronic virtual materials in the technical scheme provided by the method, the material waste is not caused and the flower arrangement practice and teaching costs are reduced.

The invention provides a cross-domain sentiment analysis method based on GCN under a lifelong learning framework, relates to the field of cross-domain sentiment classification in natural language processing, and is realized by the following process: proposing a cross-domain sentiment classification algorithm based on a graph convolutional neural network, namely CDSGCN, on the basis of proposing the CDSGCN, proposing a cross-domain sentiment classification algorithm based on a graph convolutional neural network under a lifelong learning framework, namely LLCDSGCN in combination with a lifelonglearning thought, wherein the lifelong learning is different from transfer learning or multi-task learning and other related learning tasks due to the characteristics, the limitation of isolated learning is broken through, the influence caused by time-consuming and labor-consuming manual data annotation is relieved, and the characteristics are in accordance with the initial intention of cross-domain emotion classification tasks.

The invention discloses a personalized position recommendation method based on ensemble learning, and the method comprises the steps: 1, converting a sign-in data set into a scoring matrix; 2, selecting a plurality of recommendation sub-algorithms, dividing the addresses accessed by the active users into a training sub-data set and an evaluation sub-data set, and calculating a pre-score for an address in the evaluation data set and an unaccessed address by utilizing the training sub-data set and each sub-algorithm; 3, calculating recommendation precision F1 of each sub-model by utilizing the evaluation data set, and generating a precision weight value set; 4, selecting an information gain IG as a stability index, evaluating the stability of each sub-model, and calculating a stability weight value set; 5, calculating a final total weighting coefficient for the active users, wherein the integration model fuses the pre-scores of the sub-algorithms for the unaccessed addresses according toa total weighting coefficient to generate a final prediction score; 6, evaluating the comprehensive performance of the method provided by the invention and each sub-algorithm before integration, andevaluating the effectiveness of the method.

The invention provides a through-type carrier-vehicle, which is still composed of wheels, wheel stands, a carriage and other vehicle components. The carrier-vehicle is mainly characterized in that: vertical frame tubes and base plates fixed under the center of the vehicle body are removed permanently; the vehicle is unique in modeling, safe, comfortable, convenient, multipurpose and greatly improved throughput capacity; the passenger and the cargo can enter and leave the vehicle in seconds stably; chairs, beds, tables, sheds and even boats and cases are provided at any time; the vehicle has short vehicle body, low center of gravity, high speed, flexible steering and small wind resistance, and is anti-collision and anti-skid, and does not subvert while stopping; the ground-assisting brake can be paddled for turning in site and quick backing; transporting is easy, obstacles can be passed at a high speed, and buildings and rooms can be entered easily; large pieces and ultra-large pieces can be carried; only 1/3 of the bicycle area is needed for folded storage; the operation is easy, the cost is low and the vehicle can be used without learning; the traveling and transportation efficiency is doubled; the vehicle can be used for mobile operations for companies, schools and individuals in noon break, camping, outdoor office and study; the vehicle can be used as special tools for body building, exercising, studying, rest, thinking, entertainment, traveling, processing, natural calamity fighting and first aid, and is a super tool for health preserving, body building, scientific fighting, extra development, studying, innovation, capacity cultivating and wealth and happiness creating of the whole population.