22results about How to "Fast optimization" patented technology

The invention discloses a volume video figure rendering method, system and equipment based on double-layer Gaussian splashing, a chip and a medium. The method comprises the following steps: S1, acquiring human body pictures with actions from a plurality of visual angles; s2, capturing global motion by using joint Gauss, capturing appearance details by using skin Gauss, and establishing mapping between the skin Gauss and the joint Gauss; s3, performing double-Gaussian optimization by using a coarse-to-fine optimization strategy, and ensuring time consistency; and S4, respectively providing a residual quantization compression method, a picture coding and decoding compression method and a persistent codebook compression method for different Gaussian attributes. According to the invention, through multi-view RGB video input, a double-layer Gaussian splashing and coarse-to-fine optimization strategy is used, so that the rendering quality with extremely high competitiveness is achieved; different Gaussian point cloud attributes are compressed through residual quantization compression, picture coding and decoding compression and a persistent codebook, and the method is superior to the prior art in the aspects of rendering quality, efficiency and storage efficiency.

The invention discloses a self-adaptive music synthesis sleep aiding method and system based on physiological feedback. The method comprises the following steps: acquiring physiological data of a user; preprocessing the physiological data; evaluating the sleep state of the user based on the preprocessed physiological data; defining a music parameter space according to the sleep state; optimizing music parameters based on the sleep state and the music parameter space; generating and playing music according to the optimized music parameters; while music is played, physiological reactions of the user are continuously monitored, music parameters are adjusted in real time based on the physiological reactions, and feedback circulation is formed; analyzing a long-term sleep pattern of the user based on the data obtained in the feedback loop; according to the long-term sleep mode, updating a music parameter space and an optimization strategy; security monitoring is executed, and intervention is carried out when an abnormal condition is detected; according to the invention, a plurality of key indexes such as sleep time, deep sleep proportion, sleep quality and the like are obviously improved.

The invention discloses a pelican optimization algorithm-based VMD-PCA-LSTM ultra-short-term wind power prediction method. The pelican optimization algorithm-based VMD-PCA-LSTM ultra-short-term wind power prediction method comprises a data preprocessing module, a power sequence modal decomposition module, an optimization algorithm module and a prediction model module. Firstly, abnormal data of a wind power plant are cleaned through a quartile method, missing data are interpolated, an energy difference method is adopted to select a VMD decomposition sub-modal number, and an original power sequence of complex fluctuation is decomposed; secondly, principal component extraction is carried out on high-dimensional meteorological data by adopting PCA, so that the mining difficulty of feature data is reduced; and finally, introducing a pelican optimization algorithm to carry out automatic optimization on four-dimensional hyper-parameters of a Dropout discarding rate, a learning rate, a first hidden layer unit number and a second hidden layer unit number of the LSTM prediction model so as to accelerate model convergence and improve fitting precision. The pelican optimization algorithm is simple in structure and easy to implement, has good expandability, and can optimize parameters of the LSTM model, so that the precision of wind power prediction is effectively improved.

The invention discloses a PSO-based multi-conditional constraint gas storage injection-production scheme optimization method and computer equipment, and relates to the technical field of energy storage, and the PSO-based multi-conditional constraint gas storage injection-production scheme optimization method mainly comprises the following steps: according to core data, logging data, seismic data and static fluid data of a gas storage, obtaining a PSO-based multi-conditional constraint gas storage injection-production scheme; a gas storage three-dimensional geologic model and a gas storage fluid model are constructed, fitting and inversion are performed by using a numerical simulation method and historical production data, a gas storage injection-production well is selected as an injection-production optimization object, an optimization model, a gas storage operation objective function and a multi-constraint condition are constructed, and an optimal injection-production allocation scheme is obtained by using a particle swarm algorithm. According to the multi-condition constraint gas storage injection-production scheme optimization method based on PSO and the computer equipment, the single-well injection-production gas quantity can be optimized, the gas storage production allocation scheme design is achieved, and long-term stable and safe operation of the gas storage is guaranteed.

The invention provides an operation optimization method based on a desulfurization system and a system thereof, and relates to the technical field of desulfurization equipment operation control. The health coefficient of a pump is obtained by calculating the real-time stability and the performance loss degree of a liquid circulating pump and an oxygen pump, the health degree is judged, the blockage coefficient of a spraying system is calculated, and the operation optimization of the system is achieved. The equipment failure rate of the desulfurization system is calculated and judged by setting a blockage threshold value and the health degree of a pump, historical operation parameter data is screened to form a training data set according to the cost ratio of the desulfurization system and the equipment failure rate, and an operation optimization model of the desulfurization system is constructed through a long-short-term memory neural network. The method comprises the steps that an operation optimization model is established, the operation optimization model is trained, then real-time detection parameters are collected to serve as input of the model, an optimization result is output through the optimization model, and the operation optimization model is optimized through a sparrow search algorithm and according to a loss function serving as a fitness evaluation index.

The invention discloses a two-dimensional MUSIC spectrum peak searching method based on an improved Archimedes algorithm, and the method comprises the steps: carrying out the one-dimensional MUSIC spatial spectrum function derivation of a signal model, and expanding the one-dimensional MUSIC spatial spectrum function to a two-dimensional MUSIC spatial spectrum function; secondly, taking a spatial spectrum function of the two-dimensional MUSIC as a target function of an Archimedes optimization algorithm; and then, in combination with an objective function, introducing a fusion chaos algorithm strategy to initialize a population, introducing a mutation algorithm, disturbing positions, and improving an Archimedes optimization algorithm. And finally, carrying out two-dimensional MUSIC spectral peak search through an improved Archimedes optimization algorithm, and taking the direction of arrival, namely, a pitch angle and an azimuth angle, obtained in a mode of solving maximum and minimum values through a target function as spectral peak values. The two-dimensional MUSIC spectrum peak searching method has the advantages that the optimization speed is higher, the optimization precision is higher, the robustness is better, and the two-dimensional MUSIC spectrum peak searching can be quickly and accurately carried out.

The invention provides a large-scale model pre-training method based on multivariate information perception, and the method comprises the steps: inputting collected image, text and audio data into a pre-training database, carrying out the small-batch marking of the pre-training database to form a recognition database, carrying out the pre-training of a recognition model in an unsupervised mode through the data in the pre-training database, and carrying out the recognition of the recognition model. Pre-training is performed from three perceptual states, parameters obtained by pre-training the model are used as initialization parameters of the recognition model, supervised learning training of the model is performed based on the recognition database, and a final intelligent recognition model is obtained. According to the method, on the basis of unsupervised learning, large-scale model pre-training is carried out by utilizing a discriminant model and a generative model respectively based on the powerful capability of large-scale model pre-training, so that effective utilization of massive unlabeled data is realized; the method supports the wide input of multi-modal data, and adopts mutual information maximization-contrast loss as a core optimization target, so that the model can smoothly process the complex relationship among the multi-modal data.

The present invention provides a chemical process abnormality monitoring method and a model optimization method based on a semi-supervised model, which belong to the field of chemical engineering technology. The optimization method includes: determining the loss value of the first initial semi-supervised model, the loss value is obtained by a loss function; comparing the loss value of the first initial semi-supervised model with a preset loss value; adjusting the number of hidden layers according to the comparison result of the loss value and the preset loss value to determine the range of hidden layer numbers corresponding to the semi-supervised model that satisfies the loss value less than the preset loss value; and establishing a new semi-supervised model based on the number of hidden layers in the range of the number of hidden layers, the new semi-supervised model is an optimized semi-supervised model. Through the above technical solution, the optimized semi-supervised model obtained will not have a high loss value of the semi-supervised model caused by under-fitting or over-fitting of the data, and can also be applied to chemical processes with less labeled data.

The invention discloses a method for batch equidistant adjustment of train operation plans of a CTC system based on a daily shift plan, and relates to the technical field of railway dispatching concentration. The method comprises the following steps: in response to a running line head node selected by a user, judging whether a preset condition is met; if yes, setting an adjustment range and an equidistant interval; loading the to-be-adjusted operation line meeting the condition according to the adjustment range; sorting the operation lines to be adjusted according to an increasing sequence of departure moments of the first nodes of the daily shift planning lines corresponding to the operation lines to be adjusted; synchronously traversing the sorted to-be-adjusted running lines and the corresponding daily shift planning lines, and performing batch equidistant adjustment calculation on each to-be-adjusted running line: if the departure time after equidistant adjustment is greater than the daily shift planning time, performing overall offset; otherwise, aligning to the daily shift plan time; and finally refreshing the running chart display. According to the invention, rapid and orderly adjustment of a large number of train operation plans is realized, delay propagation is effectively controlled, the workload of dispatchers is reduced, and the intelligent level of a CTC system is improved.

The invention provides a design method for optimizing a double curved surface of a building curtain wall into a single curved surface, which solves the problem that the double curved surface design of the building curtain wall is difficult to quickly and efficiently optimize into the single curved surface, and adopts the technical scheme that the method comprises the following steps of: picking up the curved surface, acquiring a double curved surface model by utilizing Rhino software, loading a Surface arithmetic unit into Grasshopper software, and calculating the single curved surface by utilizing the Surface arithmetic unit. Secondly, all curved surfaces are selected in a Rhino interface in a frame mode, a vehicle is turned back, all curved surface models in Rhino can be imported into Grasshopper, the curved surfaces are parameterized again, the unitization range of the curved surfaces is adjusted to be 0-1, and follow-up curved surface division calculation is facilitated; solving all Arc curves by utilizing an exhaustion method; comparing results by using a least square method; constructing a single curved cylindrical surface; and comparing results before and after optimization. According to the method, the design efficiency is improved, optimization parameters are unified, optimization results are compared, controllability and adjustability are achieved, a strategy is better implemented, systematic optimization is conducted on the whole design of the double curved surfaces of the curtain wall, and therefore higher efficiency and quality guarantee are brought to design and manufacturing of the curtain wall.

The invention relates to a lightweight and high-expansibility unmanned aerial vehicle model intelligent identification method. The method comprises the following steps: acquiring information of an unmanned aerial vehicle to be identified; inputting the to-be-identified unmanned aerial vehicle information into an unmanned aerial vehicle model intelligent identification model to obtain an unmanned aerial vehicle model; wherein the unmanned aerial vehicle model intelligent identification model is constructed based on a MoblieNetv3 network, and is obtained through data set training; the data set comprises real unmanned aerial vehicle signals and background signals without unmanned aerial vehicles. The method is simple, the optimization speed is high, and the unknown unmanned aerial vehicle appearing in the environment can be effectively recognized.

This application discloses a method and system for spot market bidding decisions in combined heat and power (CHP) power plants, relating to the technical field of CHP power plants. The method includes: acquiring operating data of the CHP unit and market data of the electricity spot market; constructing a heat-power coupling and coal consumption characteristic model based on the operating data and the market data, with rigid heat load supply parameters as the primary constraint; inputting the cleaned operating data and market data, along with energy conversion efficiency, energy consumption characteristics, and feasible domain of CHP output under different operating conditions, into a chaotic hierarchical gray wolf optimization algorithm to initialize the population and implement a hierarchical search strategy; defining a fitness function with profit maximization as the objective, solving for the optimal power output or optimal bid price, and guiding the CHP power plant in making bidding decisions regarding power generation or price. This application aims to solve the problems of unscientific bidding decisions and low profitability of CHP power plants in the electricity spot market.

A leaf photosynthetic rate prediction method based on an image relates to the technical field of plant photosynthesis evaluation. In order to solve the technical problems existing in the prior art, the existing photosynthetic rate measurement technology has the following problems: high cost and difficulty in large-scale application, and lack of effective analysis of the comprehensive characteristics of leaf images, resulting in low prediction accuracy. The technical solution provided by the present invention is: a leaf photosynthetic rate prediction model construction method based on an image, the method comprising: preprocessing the collected training image, including the step of dividing it into multiple intervals; counting the number of pixels in the partitions within a preset measurement area; obtaining the quantitative relationship between the number of pixels and the photosynthetic rate based on the least squares method; and constructing a photosynthetic rate prediction model based on the quantitative relationship. It can be applied to the work of analyzing the photosynthetic rate of leaves.

The invention provides a helicopter main reducer rigidity collaborative optimization method based on an orthogonal test, and belongs to the technical field of gear transmission. The method comprises the following steps: determining key parameters of the helicopter main reducer transmission system; respectively calculating the torsional rigidity, the supporting rigidity and the meshing rigidity of the transmission system; constructing a helicopter main reducer transmission system quasi-statics model based on the deformation coordination condition; in combination with orthogonal experimental design, multi-level experimental design is carried out on stiffness parameters of each level, a system uniform load coefficient is solved through a quasi-statics program, key influence factors and optimization level are determined by utilizing range analysis, and finally collaborative optimization of the system stiffness is realized. By systematically analyzing the influence of each rigidity on the uniform load coefficient of the system, collaborative optimization of the rigidity of the helicopter main speed reducer transmission system is realized, and the uniform load characteristic of the helicopter main speed reducer transmission system is improved.

The present invention discloses a method for planning an offshore wind power distributed power source based on a hybrid improved seagull algorithm, comprising: optimizing the basic seagull algorithm to obtain a hybrid improved seagull algorithm; initializing the parameters of the hybrid improved seagull algorithm; using the initialized hybrid improved seagull algorithm to solve the upper optimization model, obtain the distributed power site selection and capacity information and input it into the lower optimization model for solving, and obtain the distributed power planning information; returning the distributed power planning information to the upper optimization model; calculating whether the total cost meets the conditions, if so, outputting the distributed power planning information, if not, repeating the above steps. The present invention takes demand response into consideration during the planning process, effectively improves the amount of distributed energy consumption, reduces network losses, reduces annual comprehensive costs, realizes economical and efficient, safe and stable operation of the system, and has higher optimization accuracy and faster optimization speed in solving the problem of double-layer planning of offshore wind power distributed power sources.

The present invention provides a method for enhancing the stiffness of a lattice structure based on multi-scale topological optimization, and belongs to the technical field of structural design. First, a macroscopic model of the lattice structure is established, a unit cell model of the lattice structure is established, and the unit cell as a whole is divided into an optimized area and a non-optimized area; secondly, a displacement coupling relationship between the macroscopic model and the unit cell model is established to complete the establishment of the multi-scale model of the lattice structure; finally, an optimization solver is used to optimize the multi-scale model of the lattice structure, and the optimization result is reconstructed to obtain a full-size lattice structure model. The present invention can realize the rapid optimization of the lattice structure by establishing a multi-scale model of the lattice structure, and effectively improve the optimization efficiency of the lattice structure; by dividing the unit cell as a whole into an optimized area and a non-optimized area, any lattice structure can be designed for stiffness enhancement without establishing a theoretical model of the lattice structure, which is easy to implement.

The embodiment of the invention discloses a shop fence information dynamic adjustment method and device and a medium, and the method comprises the steps: obtaining a plurality of original user positions and a plurality of shop positions; correcting the distance between every two original user positions according to the distance between the original user positions and the co-occurrence frequency in the historical distribution behavior of the rider; clustering the plurality of original user positions according to each corrected distance, so that the original user positions with closer distances and / or more co-occurrence times are more easily clustered into one class; taking each clustering center as each clustering user position; operating an integer programming model by taking the maximization of the total order quantity of the plurality of shops as a target, and optimizing the distribution relationship between the positions of the clustering users and the positions of the shops; and generating fence information of each shop according to the optimized distribution relationship. According to the embodiment, the shop fence information can be automatically and dynamically adjusted.

The application discloses a hypoid gear digital design method and system, and is applied to the technical field of gear design and manufacturing. The method comprises the following steps: according to initial design parameters, solving the gear blank parameters of a hypoid gear pair, the forming method machining parameters of a large gear and the cutter tilt method machining parameters of a small gear; based on gear machining principles and space meshing principles, establishing the large gear tooth surface equation and the small gear tooth surface equation, determining the theoretical boundary of the rotary projection tooth surface, and dividing discrete grid nodes; constructing an objective function for iterative solving, obtaining tooth surface discrete point cloud data; constructing a cubic spline curve, reconstructing the tooth surface spline surface; establishing a gear blank solid model, cutting the gear blank by using the reconstructed tooth surface through Boolean operation, and generating a hypoid gear three-dimensional entity model. The application overcomes the disadvantages of the traditional design method, such as design and modeling simulation fragmentation, low efficiency of artificial optimization iteration, and greatly improves the design efficiency through full-automatic modeling process and fast design optimization response.

This invention provides a method for determining the optimal key furnace parameters of a gas-based vertical shaft furnace based on CFD (Computational Fluid Dynamics), relating to the field of metallurgical equipment design and numerical simulation technology. The method includes: S1, constructing a geometric model of the gas-based vertical shaft furnace and dividing it into meshes; S2, constructing a mechanistic model of gas-solid phase energy and mass flow and gas-solid interphase mass and heat transfer in the gas-based vertical shaft furnace based on Eulerian two-phase flow; S3, determining the input operating parameters of the vertical shaft furnace; S4, performing numerical simulation using the mechanistic model to solve for the chemical reaction field, temperature field, and concentration field until iterative convergence; S5, extracting key production indicators from the simulation results; and S6, determining the optimal combination of key furnace parameters for the gas-based vertical shaft furnace based on the key production indicators. This invention achieves rapid prediction of the optimal key furnace parameters for a gas-based vertical shaft furnace by establishing CFD models of gas-solid phase energy and mass flow and gas-solid interphase mass and heat transfer in a gas-based vertical shaft furnace with different key furnace parameters, performing numerical simulations, and extracting key data for comparison.

The present invention relates to the field of power grid fault location, and discloses a distributed power grid fault location method, comprising the following steps: obtaining line current information, encoding the information through a pre-set switch state code, and obtaining line state code information; constructing a switch function to calculate the line state code information, constructing a fitness function to evaluate the calculation result, and obtaining the fitness value corresponding to the line; calculating the fitness value through a pigeon flock algorithm, and determining the corresponding fault line according to the calculation result. The present invention also provides a device and system consisting of a non-transitory readable recording medium storing the above method program and a processing circuit, and the program can be called through the processing circuit to execute the above method to complete the positioning of the power grid fault. The present invention improves the speed, stability and accuracy of the pigeon flock algorithm in the optimization process, and is suitable for quickly locating faulty sections of the power grid.

The application discloses a kind of Cu-10Ni-XAl alloy of computational thermodynamics aided design, heat treatment process and application thereof, and relates to the technical field of alloy material.The process is based on copper, nickel, aluminium ternary alloy thermodynamic data, adopts CALPHAD method to draw phase diagram, designs Cu-10Ni-XAl (X=0-6wt%) alloy composition;After 1053K×180min solid solution treatment (cooling rate is greater than or equal to 10℃ / s) and 700-773K×240min aging treatment (heating rate 5-10℃ / min), high-performance alloy is obtained.The alloy strengthening phase is the dispersed distribution of L 12 Structural Ni3Al phase, when Al content is 3wt%, its Vickers hardness is 366HV, tensile strength is 674MPa, yield strength is 444MPa, elongation is 12.7%, corrosion rate is not higher than 0.02mm / a, and it is suitable for multiple fields such as ocean engineering.The application also discloses corresponding alloy, application and design system, solves the problems of traditional process depending on experience and unstable performance, and has precision and economy.

The invention relates to a multi-beam interference resource allocation method based on an improved artificial fish swarm algorithm, which comprises the following steps of: firstly, constructing a hardware interference resource model of a multi-beam jammer, and analyzing interference hardware resources required by different interference styles so as to obtain required hardware resource types; designing a solution model and constraint conditions of resource allocation according to threat level differences and interference effect differences of different interference patterns on different radars, and constructing an optimization target and a constraint function of an interference resource allocation model, thereby obtaining a multi-beam interference resource allocation model; the interference resource allocation model is solved by adopting an artificial fish swarm algorithm, and the artificial fish swarm algorithm is optimized by adopting a linearly reduced step length, mutation operation and initialization of an infeasible solution; in a word, the method has the advantages that the constructed hardware interference resource simulation effect is good, the algorithm is easy to implement, the sensitivity to an initial value is low, the earlier-stage optimization speed is high, and the global search capability is high.