93results about How to "Verify superiority" patented technology

The invention discloses a joint user access selection and resource allocation method in a cache heterogeneous network, and relates to the field of mobile communication. The method is specifically as follows: firstly, establishing a system model based on a D2D (device-to-device) of one single cell; acquiring each limitation condition and utility function Ui corresponding to the system model in a communication connection process aiming at one request user I; and then establishing a complete user access selection and resource allocation model according to each limitation condition and the utility function Ui of the system model, solving and analyzing to loosing the 0-1 variable in the model, and converting into a convex optimization problem; and finally, solving and proofing the convex optimization problem to obtain the maximum utility income of the system and a user access selection scheme under maximized resource allocation, and then performing simulation verification. The utility function is imported to jointly model, the non-convex optimization problem is converted into the convex optimization method, the system throughput is maximally improved, the micro-cellular is shunted, and the spectrum utilization efficiency is improved.

The invention provides a Doppler compensation method for synthetic aperture underwater acoustic communication and a system. With the method, a receiving end can accurately recover baseband signals sent in a diversity mode by a sending end. The method comprises steps: 101) frame synchronization detection is carried out and each frame length is obtained according to the frame synchronization detection; 102) a Doppler factor of each frame is estimated according to each frame length; 103) resampling is carried out on received signals according to the estimated value of the Doppler factor so as to finish coarse compensation of the Doppler effect; and 104) coherent superposition is carried out on signals of each diversity branch after compensation, despreading and demodulation are carried out, and the baseband signals sent by the sending end are recovered. Fine compensation is further carried out on signals after coarse compensation between the step 103 and the step 104. According to the method provided by the invention, the Doppler effect generated due to relative motion of receiving and sending nodes can be effectively resisted, coherent superposition of signals sent by each virtual subarray can be realized, gain loss of synthetic aperture processing space can be reduced and the quality of underwater acoustic communication can be improved.

The invention discloses a color image search method based on scale invariant feature transform (sift) seed region growing. The color image search method based on the sift seed region growing comprises the following steps: (1) choosing a sift seed point for an image to be detected; (2) dividing the image in to a plurality of regions by the adoption of a region growing method; (3) combining the regions according to the rule; (4) aiming at each region to extract a region characters; (5) calculating the similarity between regions of images in an image library and the region of the image to be detected; (6) structuring a weight bipartite graph; calculating the similarity between the image to be detected and the images in the image library; (7) arranging the images in the image library in descending order according to the similarity of the image to be detected. By the adoption of the mode, the color image search method based on the sift seed region growing can reduce the error problem due to inaccuracy of the images division and make up the defect that a single character search image is used.

The invention discloses a wind farm dynamic equivalent modeling method suitable for long-time domain wind speed fluctuation, which adopts quantum particle swarm optimization to optimize fuzzy C-mean clustering algorithm for clustering, A wind turbine generator set of the same group is equivalent to a wind turbine generator set (called equivalent wind turbine generator set), and the equivalent input wind speed of the equivalent wind turbine set is real-time equivalent, and the parameters of the wind turbine set are identified, and then the wind farm power collection system is equivalent, so that the wind farm dynamic equivalent model suitable for long-time wind speed fluctuation is obtained. The invention comprehensively considers the variation characteristics of wind speed of the wind farm with time, adopts the quantum particle swarm optimization fuzzy C with high clustering accuracy and strong global optimization searching ability, and adopts the quantum particle swarm optimization fuzzy C, Mean value clustering can obtain the wind farm dynamic equivalent model which can accurately reflect the external characteristics of the real wind farm.

The invention discloses a recommendation method for personalized information. The recommendation method for the personalized information comprises the steps of deciding what kind of algorithm should be adopted to generate a recommendation for the information according to the publishing time of the information; adopting a mixed recommendation algorithm of a recommendation based on content and a collaborative filtering recommendation based on a user when the interval between the browsing time and the current time is no bigger than a certain value; otherwise, directly adopting a collaborative filtering algorithm based on the user. According to the recommendation method for the personalized information, the characteristics of the information and the user can be accurately dug and described through the recommendation based on the content, and as for information, a special recommendation object, higher accuracy can be obtained through the recommendation based on the content; the information has timeliness and a hot property, the recommendation based on the content is free of a cold boot problem of new information and is free of influence of the hot degree of the information at the same time, and instead, the news content is directly dug; due to the fact that the interest of a user quickly changes along with the change of time, a more comprehensive recommendation result can be obtained by combining a result of the collaborative filtering recommendation based on the user.

The invention provides a collaborative AUV (autonomous underwater vehicle) navigation method based on robust information filtering. The collaborative AUV navigation method comprises the following steps: establishing a state equation describing an AUV navigation system and a measurement equation; estimating the state of a master AUV, and storing current time information in a state vector when a packet is transmitted to a subordinate AUV; performing information marginalization on the state vector after completion of data packet transmission; estimating the state of the subordinate AUV based on robust information filtering, receiving and processing data, performing distance measurement updating, namely performing the robust filtering process to correct distance measurement noises first and then performing the measurement updating process, and performing information marginalization on the state vector; restoring the information filtering states of the master AUV and the subordinate AUV toobtain AUV navigation information. By the collaborative AUV navigation method, the problem of low AUV positioning accuracy caused by abnormal measurement noises in underwater acoustic communication issolved, divergence of the navigation information is avoided, and the goal of high-precision real-time positioning in collaborative navigation is achieved.

The invention discloses a human face recognition method and system based on kernel non-negative matrix factorization, and a storage medium. The method comprises the steps: constructing a fractional order inner product kernel function, wherein the fractional order inner product kernel function exerts no limit on a power exponent parameter; obtaining a fractional order inner product kernel non-negative matrix factorization algorithm through the combination of the fractional order inner product kernel function and the kernel non-negative matrix factorization; and carrying out the human face recognition through the fractional order inner product kernel non-negative matrix factorization algorithm. The method solves a problem that the power parameter of a polynomial kernel function just can be an integer, and enables the selection of the power parameter to be more flexible. The method effectively irons out the changes of posture and illumination in human face recognition. Moreover, the method is quick in convergence and excellent in recognition performance.

The invention belongs to the technical field of electric power, and particularly relates to an SSA-based device. The invention discloses an LCC prediction model of a transformer substation in a high-latitude severe cold area of an LSSVM. The method comprises the following steps: setting parameters; Initializing a population; setting a fitness function; starting optimization; finishing optimization. According to the method, the global optimal solution of the LCC prediction model parameters is searched by adopting the SSA algorithm, and the obtained optimal parameter combination is used as the LS-SVM paraamters. Therefore, precision of the prediction model is enhanced. The method comprises the following steps of establishing a transformer substation LCC prediction model based on an SSA algorithm. The SSA optimization LS-is verified by performing comparative analysis on a calculation example and prediction results of other prediction models. Due to the superiority of the performance of the SVM prediction model, rapid and high-precision LCC prediction of the transformer substation can be realized when the transformer substation is newly built, and economic and technical assessment of the transformer substation in a high-latitude severe cold region is facilitated.

The invention discloses a wind power plant dynamic equivalence method combining wind field disperstiveness and unit difference, comprising the steps as follows: determining a grouping standard for a wind power plant unit according to dispersion characteristic of a wind power plant and difference characteristic of a wind turbine generator in the wind power plant; respectively dividing the wind power plant unit as different clusters according to the grouping standard; respectively adopting different wake flow models to perform equivalent modeling to the clusters of the wind power plant according to the difference of the clusters of the wind power plant; identifying equivalent generator parameter of a equivalent model so as to be equal to electric network impedance in the wind power plant. The invention has the following useful effects: the equivalence method of the invention is more rational in grouping, thereby accurately reflecting dynamic response characteristic of the wind power plant at the grid connected place, and the dynamic characteristic of the model established by the invention at the grid connected place is more close to that of the actual wind power plant.

The invention belongs to the technical field of rock stratum identification, and discloses a static penetration test rock stratum intelligent identification method and system, computer equipment and amedium. The method comprises steps: collecting data; preprocessing the data; dividing a data set; constructing a machine learning model; training a machine learning model; evaluating a classificationmodel; and using the classification model. Soil layer classification of a certain engineering base in China is taken as a research case, application of using a machine learning technology to performsoil layer division in a static sounding test is comprehensively discussed, the types of soil layers in the static sounding test are expanded, an intelligent soil layer classification technology basedon machine learning is provided for the static sounding test, a new direction is provided for research of soil layer identification in a static sounding test in China, the superiority of a machine learning technology in super-multi-class soil layer division based on the static sounding test is verified, and the feasibility of the static sounding test based on the machine learning technology in domestic use is proved.

The invention provides a multi-target resource allocation system, belonging to the field of resource allocation. The multi-target resource allocation system includes a real-time data acquisition module, a target and constraint condition generation module, a multi-target resource allocation scheme generation module, and a transmission module, the real-time data acquisition module acquires the resource processing capacity and tasks to be scheduled in each processing plant in real time, the target and constraint condition generation module determines the target and constraint conditions of the resource allocation based on the resource processing capacity and tasks to be scheduled in each processing plant acquired by the real-time data acquisition module, the multi-target resource allocation scheme generation module generates a multi-target resource allocation scheme by means of a migration bird flock optimization method, and the transmission module distributes the multi-target resource allocation scheme generated by the multi-target resource allocation scheme generation module to the processing plants to be scheduled.

The invention discloses a novel multi-modal chemical process abnormal state detection method, and aims to invent a process modeling technology capable of mining related characteristics among multiplemodals and apply the process modeling technology to multi-modal chemical process abnormal state detection. Specifically, the method implements multi-modal chemical process abnormal state detection based on the following two characteristics of multi-modal sampling data: the multi-modal sampling data are correlated; the multi-modal sampling data have respective uniqueness. Compared with the prior art, the multi-modal PCA method independently builds a model for each piece of modal sampling data, and only the uniqueness of each modal is considered; secondly, related characteristic components and independent characteristic components are respectively used for carrying out online abnormal state detection, and mined potential characteristics are comprehensively used; and finally, the superiorityof the method are compared and verified in the aspect of detecting the abnormal running state of a multi-mode continuous stirring reaction kettle through a specific embodiment.

The invention discloses a power transmission line pilot protection method based on a Kendall Tau coefficient, and relates to the technical field of power system relay protection. When an internal fault or an external fault occurs, currents on the two sides of the power transmission line MN have different characteristics. When an internal fault occurs, the current directions of the two sides of a faulty line are the same, the waveform change trends are the same, but the amplitude difference of current waveforms is large, and the current waveforms on the two sides have certain similarity and present a certain positive correlation. When an external fault occurs, the current directions of the two sides of the power transmission line MN are opposite, the current waveform change trends are opposite, the amplitudes of the current waveforms are approximately equal, and negative correlation exists between the current waveforms on the two sides. According to the obvious difference of current waveforms of faults inside and outside the line area, the Kendall Tau coefficient is used for representing and measuring the difference between characteristic sequence waveforms on the two sides of the line, a matched protection criterion of current waveforms is formed, and reliable identification of the fault line is achieved.

The invention provides a multi-target resource configuration method, and belongs to the field of resource configuration. The method comprises the steps: S1, obtaining the resource processing capability and a to-be-dispatched task in real time; S2, determining a resource configuration target and a constraint condition; S3, generating a multi-target resource configuration scheme through employing a migrating bird group optimization method; S4, transmitting the multi-target resource configuration scheme to each workshop for dispatching. The method provided by the invention can greatly improve the multi-target resource configuration efficiency.

The invention discloses a fault monitoring method based on a sequence correlation locality preserving projection algorithm, and aims to infer a brand-new data feature mining algorithm and implement fault monitoring based on the algorithm. Specifically, the method comprises the following steps: combining time sequence correlation maximization and local neighbor structure minimization into a targetfunction; in the solving process, guaranteeing the mutual orthogonal characteristic between the projection transformation vectors; and finally, executing fault monitoring by utilizing the extracted potential features and model errors. Compared with a conventional method, a sequence correlation locality preserving projection algorithm involved in the method is a brand-new feature extraction algorithm, considers the autocorrelation features and local neighbor features at the same time in the projection transformation process, guarantees the orthogonal characteristic of projection transformationvectors, and can mine the hidden useful information in the training data more comprehensively. Therefore, the method provided by the invention is a more preferable fault monitoring method.

The invention discloses a method for predicting a temperature fluctuation range of a transformer winding hotspot based on fuzzy information granulation. The method comprises the following steps: determining the number of granulation time windows according to the fuzzy information granulation quality, and extracting the effective information of original data by using the fuzzy information granulation technology; constructing a wavelet neural network prediction model based on the effective information, and optimizing the structure parameters of a wavelet neural network through a harmony search algorithm; and predicting the temperature fluctuation range of the transformer winding hotspot by using the model with the adjusted structure and parameter. The prediction accuracy of the method disclosed by the invention is relatively high, and certain guiding significance is provided for the operation and maintenance of the transformer.

The invention discloses a travel risk control method in a high-risk road segment. The method comprises the steps that 1, a road segment in unit length is selected, and the high-risk road segment in anaccident set is obtained through calculation; 2, traffic flow on the high-risk road segment is divided into two states; 3, a high-risk region is divided into two functional segments, namely an early-warning segment and the high-risk road segment; and 4, according to a spatial layout of the early-warning segment and the high-risk road segment shown in the step 3, the arrangement position of information publishing equipment is obtained. According to the method, firstly, the appropriate road segment unit length is selected for a research object, several different combinations are divided in combination with a dynamic segmented movement step length method; next, a Relim algorithm is adopted to mine a frequent item set in each combination, the number of traffic accidents and casualties are quantitatively processed, and the traffic accident high-risk road segment small in length and in the accident set is obtained; and then accident influence factors in the found high-risk road segment aresubjected to association rule mining.

The invention belongs to the technical field of radars, and discloses a high-frequency ground wave radar working frequency optimization method and system, a storage medium and application, and the method comprises the steps: obtaining frequency spectrum data, employing a two-dimensional OS algorithm and a double threshold algorithm to process the frequency spectrum data, and obtaining the frequency spectrum information of anti-impact interference processing; constructing an LSTM network and carrying out model training and prediction deviation evaluation; and selecting a quadratic curve trapezoidal window, sliding the window function on the whole frequency spectrum to extract the average power spectrum and fluctuation degree information in the window, calculating the total score of the sliding window, and further selecting the optimal working frequency according to the score. The real-time performance and the accuracy of frequency selection are enhanced through the improvement of a frequency spectrum prediction scheme in the invention; the time sequence is predicted through the LSTM model, frequency spectrum information at the current moment is trained and learned instead of simple data frequency shift, the prediction is closer to a real value, reliability is greatly enhanced, sliding window frequency selection is carried out through the quadratic curve trapezoidal window, and an interference frequency band can be effectively avoided.

The invention discloses an electrical internet distributed low-carbon economic dispatching method with demand response. The method comprises the following steps of establishing a stepped carbon transaction cost model corresponding to carbon transaction cost, establishing a P2G carbon transaction cost model corresponding to the earnings obtained by the carbon emission permit transaction market, according to the stepped carbon transaction cost model and the P2G carbon transaction cost model, establishing a comprehensive carbon transaction cost model to obtain a comprehensive carbon transaction cost, adjusting the load capacity according to the excitation type demand response to construct an IDR-DSM model, adjusting the load capacity according to the price type demand response to establish a PDR-DSM model, establishing a centralized low-carbon economic dispatching model representing the total operation cost on the premise of meeting the power grid operation constraint condition, the natural gas operation constraint condition and the power-gas coupling constraint condition, on the basis of the centralized low-carbon economic dispatching model, adopting an S-ADMM algorithm to solve and establish a completely distributed dispatching model, and ensuring the low-carbon economy of the network.

The invention discloses a medical electronic nose gas concentrating system, and a method. The medical electronic nose gas concentrating system comprises a flow rate control unit, a temperature control unit, and an adsorption tube containing a mixed adsorbent. The mixed adsorbent is selected from CarbopackTMB and Tenax GR based on the adsorption capacity of gases; OD1 / 4inch, 7inch Length standard adsorption tube is adopted; and the adsorption tube contains the mixed adsorbent with a thickness ranging from 120 to 150mm; and a first grade Peltier cooling system is adopted by the mixed adsorbent. A gas concentration system is adopted by the medical electronic nose gas concentrating system, so that the detection capacity and capacity of resisting disturbance of the medical electronic nose with the gas concentrating system are both improved, a problem of the prior art that electronic noses can not be directly applied to respiratory gas disease diagnosis because of poor gas sensor detectability is solved, development of olfactory analogue technology of the state is promoted, the gap between the prior art and international advanced level is narrowed, disease early diagnosis is realized, and great practical significance is achieved.

The invention discloses a method for online measuring an inclined axis hole by a five-axis machine tool, and relates to the technical field of measuring methods of machine tool parameters. The method includes the steps: performing modeling on multi-body system kinematic theories and errors; analyzing five-axis machine tool kinematic errors; optimizing machine tool probe calibration errors; optimizing a geometric error evaluation method of online measurement of the inclined axis hole. The method is simple in theory, easy to programme, conveniently applied in practical engineering and applicable to axis solution and cylindricity error evaluation of the inclined axis hole. Specific experiments verify superiority and feasibility of online measurement of the five-axis machine tool.

The invention relates to the technical field of distributed radar systems, and discloses a moving target positioning method for a distributed MIMO radar system, which takes a bistatic distance (BR) asan observed quantity, and comprises a distributed MIMO radar positioning scene, and constraining overall least square positioning; and the method comprises the following steps: constructing an auxiliary vector, performing linearization processing on an observation equation of a nonlinear bistatic distance, establishing a positioning problem as a constrained total least square model, and solving the model by adopting a Newton iteration method, thereby obtaining target position estimation. According to the method, BR observation is utilized to estimate the position and the speed of the target,no additional requirements are required for the number of the transmitting units and the number of the receiving units, errors existing in an equation coefficient matrix and a data vector are avoided,and compared with an existing BR positioning method, the method has remarkable advantages and superiority in the aspect of positioning precision when the measurement error is large.

The invention discloses a non-gaussian process monitoring method based on novel dynamic independent component analysis, and aims to combine the advantages of a dynamic internal principal component analysis model which can deal with autocorrelation dynamic data with an independent component analysis model which can deal with non-gaussian data. Specifically, the non-gaussian process monitoring method includes the steps that firstly, a dynamic internal principal component analysis algorithm is used for correspondingly extracting autocorrelation dynamic characteristic components and cross-correlated static characteristic components; secondly, after whitening processing of the characteristic components, combined whitening characteristic components are used as initial independent components anda dynamic independent component variable model is obtained iteratively; and finally, based on the dynamic independent component variable model, dynamic non-gaussian process monitoring is implemented.In conclusion, according to the non-gaussian process monitoring method based on novel dynamic independent component analysis, the ability of the dynamic internal principal component analysis algorithmseparately extracting dynamic components and static components is utilized, and an independent component analysis algorithm capable of extracting non-gaussian characteristic components is further combined. Therefore, the non-gaussian process monitoring method based on novel dynamic independent component analysis is the feasible dynamic non-gaussian process monitoring method.

The invention relates to a vehicle driving cost evaluation method based on a data driving scene, and belongs to the field of new energy vehicles. The method comprises the steps of obtaining driving historical data in a specific area, performing data preprocessing, and performing working condition fragment division; carrying out dimensionality reduction on multi-dimensional characteristic parameters of working condition segments based on principal component analysis; utilizing an IABC-Kmeans algorithm to establish a typical working condition feature set, reflecting a synthetic working condition of a specific driving style and a driving habit in a certain area according to the quantitative proportion of each clustering sample and correlation recombination of samples in the class and clustering center parameters, and carrying out comparison verification of statistical features on the synthetic working condition and an original data set; designing a unified quantification method for battery aging, fuel consumption and electric quantity maintenance, and establishing a DDPG multi-target energy management optimization model integrated with expert experience, so that the strategy has higher training efficiency on the premise of ensuring optimality. According to the invention, a reference can be provided for a more accurate vehicle driving cost evaluation method.

According to the whale optimization algorithm-based smooth path planning method provided by the invention, the path length, the curvature and the curvature derivative in an improved WOA optimization curve path are adopted, and a good result is obtained. In the improved WOA algorithm, a Levy flight strategy is introduced to add local disturbance to the algorithm, the diversity of solutions is increased, and local optimum is avoided. And meanwhile, a fractional order expansion equation is introduced, so that each generation of whale individuals is influenced by the positions of the previous generations, and the search capability of the algorithm is further enhanced. In 23 reference functions, the performance of the improved WOA algorithm and other algorithms is tested and compared, and the superiority of the improved WOA algorithm is verified. Finally, analog simulation experiments are carried out in two different maps, a high-order Bezier curve is adopted in a smoothing method, and continuity of the path curvature is guaranteed. Compared with the existing other algorithms, the improved WOA algorithm has the advantages that the performance is obviously improved, and the operation time is shorter.

The invention discloses a rectification process real-time monitoring method based on on-line sampling data driving, and aims to drive characteristic transformation in real time according to on-line sampling data so as to realize real-time monitoring of the operation state of a rectification tower by using most representative characteristics. Specifically, the method does not use a fixed projectiontransformation vector to extract features, but uses online data to drive feature analysis and extraction in real time. Compared with the traditional method, the method provided by the invention is very simple and clear to implement, and almost has no off-line modeling stage. The offline modeling stage mainly relates to standardization processing and basis matrix calculation. Secondly, according to the method, aiming at each piece of online measurement sample data, a projection transformation vector which can most distinguish the online measurement sample data from normal working condition data is searched. From this point, the characteristic components extracted by the method are most beneficial to monitoring the fault data. And finally, the superiority of the method in real-time monitoring of the rectifying tower process is verified through a specific embodiment.

The invention relates to the field of wave absorbers, and concretely relates to an adjustable ultra-wideband terahertz absorber based on metal and graphene. The absorber comprises an ideal electric conductor layer, a dielectric layer and a bimetallic ring which are sequentially arranged from bottom to top, a graphene layer is arranged in the dielectric layer, and the bimetallic ring is composed of a first metal ring and a second metal ring which are concentric. The surface conductivity of the graphene changes along with the Fermi level, and when the chemical potential of the graphene is changed, the broadband adjustability can be realized, so that the purpose of tunable absorption is achieved; and the bimetallic ring and the graphene layer are combined to design the metamaterial wave absorber, so that the metamaterial wave absorber not only has the high absorptivity of a metal metamaterial wave absorber, but also has the regulation characteristic of a graphene metamaterial, and has the characteristic of broadband high absorption. The structure is simple, and electromagnetic wave absorption of other wave bands can be realized through scale transformation of the size.

The invention discloses a wind power plant dynamic equivalence method combining wind field disperstiveness and unit difference, comprising the steps as follows: determining a grouping standard for a wind power plant unit according to dispersion characteristic of a wind power plant and difference characteristic of a wind turbine generator in the wind power plant; respectively dividing the wind power plant unit as different clusters according to the grouping standard; respectively adopting different wake flow models to perform equivalent modeling to the clusters of the wind power plant according to the difference of the clusters of the wind power plant; identifying equivalent generator parameter of a equivalent model so as to be equal to electric network impedance in the wind power plant. The invention has the following useful effects: the equivalence method of the invention is more rational in grouping, thereby accurately reflecting dynamic response characteristic of the wind power plant at the grid connected place, and the dynamic characteristic of the model established by the invention at the grid connected place is more close to that of the actual wind power plant.

The invention discloses an order-perceived remote sensing image instance increment detection method. The method comprises the following steps: acquiring a remote sensing image instance; according to whether there is a real label, designing an order score calculation function to respectively evaluate uncertainty and inaccuracy of an inference result of the target detection model, and measuring an incremental learning value in a newly added instance and an incremental learning value in a historical instance; while the target detection model is finely tuned, training loss is weighted based on order scores to balance the contribution of the training data of the value difference. According to the method, streaming data with different values can be continuously learned, and data differences are comprehensively concerned; newly added instances and historical instances are sorted in a self-adaptive mode, and a unified sample weighting direction is provided for model training.

The invention discloses a photovoltaic power generation power prediction method based on a long short-term memory neural network. The method comprises: obtaining historical data of photovoltaic power generation power, dividing the data into a training set and a test set by using a cross validation method, and performing normalization processing; substituting the data set into a long short-term memory neural network quantile regression model to obtain the predicted power of photovoltaic power generation; and finally, verifying a prediction result by adopting a root mean square error method. According to the method, the mode of the probability density function is used as a point prediction result, so that the prediction precision can be further improved, and the actual value of the prediction point has a relatively large probability to appear at the mode of the wind power probability density function; and the interval coverage rate meets the expected requirement, the interval average width is minimum, the quantile is minimum, and the superiority of the model in interval prediction compared with a traditional model is verified.