79results about How to "Non-linear" patented technology

In conducting laser annealing using a CW laser or a quasi-CW laser, productivity is not high as compared with an excimer laser and thus, it is necessary to further enhance productivity. According to the present invention, a fundamental wave is used without putting laser light into a non linear optical element, and laser annealing is conducted by irradiating a semiconductor thin film with pulsed laser light having a high repetition rate. A laser oscillator having a high output power can be used for laser annealing, since a non linear optical element is not used and thus light is not converted to a harmonic. Therefore, the width of a region having large grain crystals that is formed by scanning once can be increased, and thus the productivity can be enhanced dramatically.

The invention discloses a trajectory tracking sliding mode control method for a spraying mobile robot. The method comprises the following steps of: performing mechanism analysis on a mobile robot, and establishing a mobile robot kinematic model with non-integrity constraint; establishing a controlled object mathematical model of each branch controller of a wheeled mobile robot provided with a motor driving shaft disturbance term; identifying a traveling path by utilizing a computer vision system, and determining an expected motion track of each branch driving motor according to the kinematic model deduced in the previous step; detecting the rotating speed of the motor, calculating the actual motion angular velocity and actual motion angular acceleration of left and right driving motors of the mobile robot, and calculating the deviation and deviation derivative between the expected angular velocity and the actual angular velocity of each driving motor; establishing a sliding mode switching function which meets the speed control requirement of the driving motor; determining the sliding mode controller control quantity of the left and right driving motors of the mobile robot on the basis of the sliding mode surface function s; and respectively transmitting the control quantity of the motor of the mobile robot to the left and right driving motors.

This invention relates to a single wheel robot system and its control method. The robot is an intelligent self-control and thus self-balancing unicycle riding robot. The control method is the balance control method of the static imbalance unicycle robot. The single wheel robot includes mechanical body and control system; the body contains a single wheel in the substructure which can rotate around for balance; the control system comprises state sensors, motion controller, servo-driven controllers, and a power system. Among them, the motion controller receive signals from the state sensors, in accordance with control procedures for processing of the received signal, thereby issuing control instructions. The servo drive controller receives the control instructions and controls the motors of the robot to adjust posture to be balanced.

The invention discloses a deep belief network image recognition method based on Bayesian regularization and belongs to the field of artificial intelligence and machine learning. The deep belief network plays a more and more important role in the field of digital detection and image recognition. The invention provides a deep belief network based on Bayesian regularization on the basis of the network sparsity characteristic and changes of connection weights to solve the problem of overfitting in the training process of the deep belief network. By applying Bayesian regularization to the network training process, balance between error decreasing and weight increasing is effectively adjusted. The classification experiment of a digital script database proves effectiveness of the improved algorithm. An experimental result shows that in the deep belief network, the deep belief network image recognition method can effectively overcome the overfitting phenomenon and improve accuracy of digital recognition.

In allusion to the characteristics that in a sewage treatment process, the effluent water quality cannot reach the standard, the energy consumption is relatively high and the like, the invention provides a multi-objective real-time optimization control method for the sewage treatment process, by which optimization control over concentrations of dissolved oxygen SO and nitrate nitrogen SNO are achieved in the sewage treatment process. According to the optimization control method, an established energy consumption model and an established effluent water quality model, which are based on radial basis functions, are used as optimized objective functions, the optimized objective functions are treated via a multi-objective particle swarm algorithm to obtain optimized set values of the dissolved oxygen SO and the nitrate nitrogen SNO, and tracking control is performed on the optimized set values of the dissolved oxygen SO and the nitrate nitrogen SNO through a fuzzy neural network. According to the optimization control method, the problem of multi-objective real-time optimization control over the sewage treatment process is solved, the energy consumption is reduced on the basis that the effluent water quality is ensured, and high-efficiency and stable operation of a sewage treatment plant is promoted.

The invention relates to the field of linear motor control, in particular to a method for restraining the pulsation of a thrust system of a linear motor; the method adopts a sliding mode thrust controller and a magnetic flux controller, the thrust deviation and the magnetic flux deviation are selected as controlled variables, an integral sliding mode surface which is composed of the thrust deviation and the magnetic flux deviation design the sliding mode motion trajectory, thereby a system can move according to the sliding mode trajectory and the output thrust and the magnetic flux can better track specified value, wherein, a fuzzy controller utilizes the fuzzy control method which uses language information and data information to approach any specified continuous function to solve the jitter problem of a sliding mode control system. The method has the advantages that: a direct drive control system of the linear motor has strong coupling property, nonlinear property, multiple variables and unique end effect, the use of the direct thrust control method of the fuzzy sliding mode linear motor with high robust performance can solve the impacts of various factors on the control performance, thereby achieving the purposes of restraining the thrust pulsation of the linear motor and obtaining great dynamic response performance.

The present invention discloses a method for controlling rotating speed synchronization of a dual-permanent magnet synchronous motor drive system. The method comprises the steps of establishing a discrete mathematical model of a permanent magnet synchronous motor, wherein the discrete mathematical model comprises a voltage equation and an electromagnetic torque equation of the permanent magnet synchronous motor in a d-q axis coordinate system, and a motion equation of the permanent magnet synchronous motor; according to a slide mode control principle, designing rotating speed ring controllers of two permanent magnet synchronous motors into integral slide mode speed controllers; and according to a cross coupling principle, designing a speed synchronous controller, and compensating current rings of the two permanent magnet synchronous motors. The system has good robustness and is rapid when load disturbance occurs, and the rotating speed tracking and synchronization performance of the dual-motor drive system are effectively improved.

The invention discloses a water bloom prediction method based on a space-time sequence hybrid model, and belongs to the technical field of water environment prediction. The water bloom prediction method comprises the following steps: firstly, extracting a large-scale nonlinear trend term of water bloom spatio-temporal data based on a deep belief network; establishing a space weight matrix based onthe geographic positions of the multivariate space-time meteorological monitoring points; then extracting a small-scale residual term and carrying out modeling again; superposing the large-scale nonlinear trend term prediction value and the small-scale residual term prediction value, and obtaining a meteorological prediction value of the target water area according to an inverse distance weighteddifference method; and using ANFIS fusion to predict the water quality and meteorological data of the target water area. According to the method, the number of influence factors of water bloom outbreak is increased, so that the result of water bloom modeling prediction is more accurate, and the influence effect of the surrounding water area on the target water area can be reflected more truly. The method is high in applicability, can be used under the condition of bloom space-time sequence data of different water areas, is suitable for predicting bloom outbreak under different water qualitiesand weather conditions, and has universal applicability.

The invention provides a method for controlling content ranges of components in a rare earth extraction and separation process. The method comprises the steps of establishing an echo state network model of the rare earth extraction process based on the characteristics of the flow rate / component content process control of the rare earth extraction and separation process; and putting forward a method for controlling the content ranges of multiple components in rare earth extraction and separation by generalized prediction to realize content range control of multiple components in rare earth extraction and separation. The traditional method adopts a soft measurement model (a static model) in extraction process equilibrium state, which cannot realize the online prediction of the contents of the components in the extraction process easily and cannot establish a precise control model easily, so as to affect the tracking control effect of the rare earth component contents. According to the control method provided by the invention, adjustment is implemented according to the range control strategy, and the calculation is optimized to obtain the accurate control amount of the rare earth extraction process, so that the component content of the rare earth extraction process meets the range control requirements, and the quality of the products at both ends is ensured. The method provided by the invention is suitable for modeling and optimizing control of the rare earth extraction process.

The invention provides a thickener underflow concentration prediction method based on integrated learning, and belongs to the technical field of mining. The method comprises the following steps: obtaining actual production historical record data, storing the actual production historical record data in an enterprise database, then preprocessing the obtained data set, and constructing a training setand a test set by using the preprocessed data; and an integrated learning method is adopted, the constructed training set and test set are utilized to establish a model, accurate prediction of the underflow concentration of the deep cone thickener is realized, and finally, a prediction result is displayed through a visual platform. According to the method, most factors influencing the underflow concentration can be comprehensively considered, so that the bottleneck problem of insufficient one-sided consideration when an existing underflow concentration prediction model considers the influencefactors is solved. And an integrated learning model is used, so that the problems that a single machine learning model is limited in learning capability and large-scale data cannot be processed are solved, and more effective and accurate reference is provided for control of the thickener.

The invention discloses a dissolved oxygen control method based on a fuzzy neural network. The method comprises the following steps: S1, modeling is carried out; S2, fuzzy identification is carried out, and a fuzzy identification method is adopted to identify the structure and parameters of an object model; S3, a clustering method is adopted to determine the number of rules; S4, a neural network is used for learning reasoning data of an expert, a fuzzy reasoning rule is acquired, and each connection weight is adjusted at the same time; S5, after a least square method is used for order identification, a transitive closure method is used for clustering analysis according to a fuzzy similarity relation; S6, a Smith predictor is introduced; and S7, two independent control circuits of blower pressure and oxygen dissolution are adopted. The wastewater treatment effects are good.

The invention relates to an outlet ammonia nitrogen concentration prediction method based on an adaptive recurrent fuzzy neural network, which belongs to the control field and the water treatment field. As that measure process of ammonia nitrogen concentration in the effluent of the current sewage treatment proces is tedious, As that cost of instrument and equipment is high, the reliability and accuracy of the measurement result are low, and the like, the problem that the concentration of ammonia nitrogen in the effluent is difficult to be measured is sol by utilizing an adaptive recurrent fuzzy neural network to realize the prediction of the concentration of ammonia nitrogen, which is a key water quality parameter, based on the biochemical reaction characteristic of the municipal sewage treatment; The results show that the recurrent fuzzy neural network can be fast. Accurately predicting the concentration of ammonia nitrogen in wastewater treatment effluent is helpful to improve the concentration and quality monitoring level of ammonia nitrogen in wastewater treatment process and strengthen the fine management of municipal wastewater treatment plant.

The invention discloses an electroencephalogram multi-domain feature extraction method based on multivariate variational mode decomposition. The method comprises the following steps: firstly, carryingout adaptive decomposition on original electroencephalogram multi-channel data by using multivariate variational mode decomposition (MVMD), and then extracting time domain features and nonlinear dynamics features of signals from intrinsic mode function (IMF) components obtained by decomposition; meanwhile, combining the IMF components to construct a new signal matrix, extracting spatial featuresof the reconstructed signals by adopting a common spatial pattern (CSP) method, and combining the time domain, nonlinear dynamics and spatial domain features; and finally classifying the feature set through a support vector machine (SVM). According to the method, important information components related to a specific task can be effectively distinguished, and a new idea is provided for feature extraction of the electroencephalogram signals.

A low-frequency and large-displacement angular vibration table comprises a case, a work tabletop, a main shaft driving the work tabletop to rotate, a moving coil assembly, a magnetic circuit assembly, a motor driving the magnetic circuit assembly to rotate, a closed-loop control assembly of the motor, an electric viscoelastic feedback control assembly, an air bearing and an angular displacement sensor. The main shaft is fixedly connected with the moving coil assembly, and the magnetic circuit assembly is fixedly connected with a rotor of the motor through a connector. The moving coil assembly comprises a moving coil base body and coils, wherein the moving coil base body is fixedly connected with the main shaft. The magnetic circuit assembly comprises a magnetism guide ring, a central magnetic pole and magnetic steel, wherein the magnetism guide ring, the central magnetic pole, the magnetic steel and an air gap form a closed magnetic loop. The central magnetic pole is located in the magnetism guide ring, the magnetism guide ring is coaxial with the central magnetic pole, and the magnetic steel is located between the magnetism guide ring and the central magnetic pole and attracted to the central magnetic pole. The moving coil assembly is located between the magnetic steel and the magnetism guide ring and is coaxial with the magnetism guide ring. The low-frequency and large-displacement angular vibration table has the advantages that the distortion factor of output wave forms is small, and the output angular displacement is large.

The invention discloses a hyper-spectral classification method based on low-order mutual information and spectral context band selection, and mainly aims to solve the problem that the computational complexity is high and the classification performance is poor for hyper-spectral image classification in the prior art. The implementation scheme comprises the following steps: first, automatically removing noise bands based on context priori information between neighborhoods of spectral bands of hyper-spectral images; second, selecting a band subset with low redundancy and high information content from a band set obtained after removal of the noise bands by a sequence forward search method according to self information of a maximum single band and regularization mutual information between a minimum band and other bands; and finally, performing object classification by using the selected bands. By adopting the method of the invention, a high-accuracy high-efficiency result of hyper-spectral image classification can be obtained. The method can be used for object distinguishing and identifying in soil survey, urban environment monitoring, disaster evaluation and other fields.

The invention relates to an electromyography signal noise reducing and aliasing removing method based on combination of second-generation wavelet transform and ICA and aims to overcome some revealed defects when the current signal aliasing removing methods are applied to electromyography signals. The method includes: a second-generation wavelet noise reducing algorithm is first used to filter noises in the electromyography signals, the second-generation wavelet noise reducing algorithm is applied to the electromyography signals after noise reduction is smooth, unnecessary vibration in waveforms is restrained, and the electromyography signal features are evident; ICA separation is performed on the vague signals after noise reduction so as to fast remove the aliasing components in the signals. The method has the advantages that by the pre-treatment, interference in the signals can be removed greatly, and convenience is brought to subsequent researches such as electromyography signal feature extraction and action identification.

The invention discloses synchronous tracking and controlling method for a motor train unit based on a distributed model, which aims at the structural characteristic that the motive power distributed motor train unit is composed of a plurality of mutually coupled traction / brake control units. The method is based on motor train unit running process data and a traction / brake characteristic curve, utilizes modeling approach of data driving to build a motor train unit running process distributed model, and adopts a subspace method to identify, so as to obtain a model parameter. The invention provides the synchronous tracking and controlling method for the motor train unit based on the distributed model, so as to obtain an accurate control quality, and allow the running speed of each unit of the high-speed motor train unit to track the target curve accurately, achieve high-precision synchronous tracking control for each unit during high-speed motor train unit running process, and guarantee the motor train unit to run on time, safely, and efficiently. The method is simple and practical, and is suitable for high-speed motor train unit running process modeling and synchronous tracking control.

The invention discloses a metal corrugated-metal rubber composite material and a preparation method thereof. A matrix metal corrugated material and stamped type metal rubber are adhered together through adhesives to obtain a multifunctional composite porous material, and preparation of the metal corrugated-metal rubber composite material is realized; the metal corrugated-metal rubber composite material has wide application prospect in fields of transportation, mechanical manufacture and military, structural damping can be effectively increased, nonlinear and variable damping properties are achieved, earthquake, vibration and sound absorption of the structure can be improved, and high strength, high rigidity and excellent damping performance in machine-building industry can be exerted; in addition, the composite porous composite material has the advantages of total environmental adaptability, corrosion resistance, high-and low-temperature resistance, no ageing, vacuum non-volatilization and the like and is light, simple in manufacture, low in cost and long in service life.

The invention discloses a low-voltage power distribution system series fault arc identification method based on all-phase deep learning. In an existing low-voltage power distribution network fault, the identification method for a series fault arc is easily disturbed by a noise and spectrum leakage, an identification effect is affected, identification efficiency is not high, and stability is not high either. In the invention, the above problems are solved. The method comprises the following steps of under a low-voltage alternating current system, carrying out current signal collection on different loads in a low-voltage loop; carrying out all-phase discrete Fourier transform on a collected current signal, carrying out full-phase spectrum characteristic quantity extraction of a load, and constructing an all-phase spectrum characteristic vector; constructing a deep learning neural network model based on Logistic regression, carrying out deep learning training on all-phase spectrum characteristic quantity under the different loads and different operating states till that the model converges; and using the trained model to complete screening of different load types and identification ofwhether the series fault arc occurs.

The invention discloses an MBR membrane permeable rate intelligent detection method based on a recursion RBF neural network and belongs to the field of sewage processing water quality parameter online detection. In an MBR membrane sewage processing process, the problem of pollution affects the outlet water quality of a membrane and the life of the membrane and prevents large scale application of the membrane; and the MBR membrane sewage processing process is severe in random interference and also has the disadvantages of high nonlinearity, large time variation and severe lag, and the pollution cannot be directly measured and detected in an online mode. According to the method based on feature extraction, six types of process variables highly relevant to a permeable rate are obtained; and at the same time, by taking the membrane permeable rate as output of a model and the six types of process variables as input of the model, a soft measurement model of the membrane permeable rate is established based on the recursion RBF neural network, and real-time detection of a membrane pollution degree is completed, quite good precision is obtained, a result indicates that the permeable rate can be rapidly and accurately predicted, stable and safe operation of the MBR membrane sewage processing process is ensured, and the quality and the efficiency of membrane sewage processing quality are improved.

The invention provides a sewage treatment process optimization control method based on multiple gradient descent, focuses on the disadvantages that the linearity, coupling and uncertainty are high in the sewage treatment process and aims to control the concentrations of dissolved oxygen DO and nitrate nitrogen SNO simultaneously in the sewage treatment process. The control method includes establishing a multiple objective function of a control system, the multiple objective problem of the sewage treatment process is solved by the optimization method based on multiple gradient descent, and the purposes of controlling the concentrations of dissolved oxygen DO and nitrate nitrogen SNO by controlling the aeration rate and the internal recirculation flow after the optimization are achieved; the problem of the multiple objective problem of the sewage treatment process is solved, the control precision of the dissolved oxygen DO and nitrate nitrogen SNO in the sewage treatment process is improved, normal operation of the sewage treatment process is guaranteed, and the efficient and stable operation of a sewage treatment plant is guaranteed.

The invention discloses a decoupling control method of rare earth extraction process. The decoupling control method of rare earth extraction process provides a kernel function extreme learning machinecomponent content model, directing at the characteristics of multivariable, strong coupling and nonlinearity in the rare earth extraction process group, in allusion to the data characteristics of element component content, extractant flow and washing agent flow of monitoring points at two ends of the rare earth extraction process, and establishes a multi-input and multi-output model of rare earthextraction process and converts the multi-input and multi-output model into a plurality of multi-input single-output sub models, by combining with the dynamic process data of different operating phases of the rare earth element CePr / Nd extraction process. The decoupling control method of rare earth extraction process adopts a strategy of performing adaptive adjustment on the deviation weight in the system performance index according to the deviation between the reference trajectory value and the model output value, in the control loop, so as to design a decoupling controller of the rare earthextraction process, thus reducing the coupling between each control loop to realize approximate decoupling control to guarantee the quality of the export products at both ends.

The invention discloses a glass-glass combined light waveguide, comprising glass substrate with light guiding region and functional glass substrate interlinked, where the functional glass substrate has light limit region and the light guide region and the light limit region together compose core part of the combined light waveguide; even if on the condition that the refractivity of the functional glass substrate is higher than that of the light guide region, by the action of the light limit region, the light transmitted in the light waveguide can not be transmitted in the form of radiation mode. In this case, it assures interaction between the transmitted light and the functional glass substrate and has simple making process; the functional glass substrate has light amplification, nonlinear, magneto-optical or electrooptical property, and this light waveguide structure makes full use of the functions of the functional glass substrate. And the combined waveguide can integrate different functions into the same light waveguide component, implementing miniaturized and multifunctional light integration devices.

The invention discloses an elastic return-difference-free 3Z planetary reducer. The elastic return-difference-free 3Z planetary reducer comprises a central sun wheel, planetary wheels, planetary wheel shafts, a planetary frame, a fixed inner gear, a rotating inner gear and a bearing and is characterized in that a layer of elastic material is additionally arranged at the matching position of each planetary wheel shaft and a hole, a layer of elastic material is additionally arranged at the matching position of a sun wheel shaft and a hole, one planetary wheel is adjusted to enable the central distance between the planetary wheel and the sun wheel to be smaller than theoretical value, the central sun wheel is eccentrically placed on a shell of the fixed inner gear, the corresponding central distance between each planetary wheel and the fixed inner gear is larger than theoretical value, through elastic material deformation, the central sun wheel, the corresponding planetary wheels, the fixed inner gear and the rotating inner gear are tightly engaged, and no side gaps exist.

The invention relates to an intelligent detection method for the water permeability of an MBR membrane based on the deep belief network (DBN), and the method comprises the steps: (1), determining a target variable and a characteristic variable; (2), designing a soft measurement model of membrane water permeability, and using the DBN to establish a soft measurement model for predicting the water permeability of the membrane; (3), correcting the built soft measurement model for predicting the water permeability of the membrane, and obtaining a simulation error graph and a prediction result graph; (4), predicting the water permeability. Based on the DBN, the soft measurement model for predicting the water permeability of the membrane is established, which reduces the computational complexityof water permeability, realizes on-line accurate measurement and real-time correction of water permeability, and provides an effective method for predicting the pollution state of the membrane in wastewater treatment, and improves the working efficiency and economic benefits of the MBR membrane wastewater treatment process.

The invention discloses a material for preparation of a zinc oxide piezoresistor. The material comprises, by weight, 93 to 96 parts of zinc oxide, 0.05 to 1.3 parts of tin oxide, 1.5 to 1.8 parts of bismuth trioxide, 46 to 49 parts of yttrium oxide, 46 to 49 parts of ytterbium oxide and 9 to 12 parts of silicon dioxide. According to the invention, yttrium ions and ytterbium ions are introduced andbismuth trioxide is used, so the performance of the zinc oxide pressure-sensitive material is improved.

The invention discloses a fee-paying behavior analysis method based on the SOM neural network clustering algorithm. The method comprises steps that data of all the fee-paying user basic attribute information and the fee-paying habit attribute information of an entire region is acquired to form a data set; behavior index parameters, the customer classification quantity and connection right constraint conditions of the data set are determined, and the SOM neural network is constructed; a part of samples in the data set is selected, training of each learning mode of the SOM neural network is sequentially carried out, and each of the connection weights connected with the winning neuron is continuously optimized and corrected until the correction amount satisfies the set value; the data set isclassified through utilizing the optimized SOM neural network to acquire target behavior index parameters, the classification result of the customer classification quantity is satisfied, an average value of each index of the data set is calculated, and the fee-paying behavior clustering result is acquired. The method is advantaged in that certain relevance among influence factors can be acquired through data mining, and classification and further research of client fee-paying behaviors are further facilitated.

The invention discloses a caterpillar track type multifunctional emergency service vehicle and belongs to the field of emergency rescue and disaster relief equipment. The caterpillar track type multifunctional emergency service vehicle comprises a chassis. A traveling device is installed under the chassis. The traveling device comprises six pairs of load-bearing wheels installed under the chassis, two pairs of driving wheels, caterpillar tracks and track supporting wheels used for supporting the caterpillar tracks. A suspension device is installed between each pair of load-bearing wheels and the chassis. A winch is fixed to the front end of the chassis, and a cab and a rotary table are fixed to the chassis. A multifunctional arm and an operation room for controlling the multifunctional arm are installed on the rotary table. A quick-change connector is installed at the end of the multifunctional arm. The caterpillar track type multifunctional emergency service vehicle has the advantages that the multifunctional arm can be selectively connected with an excavation bucket, a water suction pump, an extension jib, a rapid sealing clamp and the like rapidly, so that the vehicle is multifunctional; the vehicle is high in adaptability to various terrains, suspension locking and leveling of the vehicle body are achieved, and the vehicle is good in running stability and comfort level; and when the vehicle is stuck in a bog and sludge and cannot go forward, the winch is used for self-rescue.