16 results about "Stochastic field" patented technology

The invention relates to the technical field of data center optimization, and particularly discloses a data center two-stage stochastic optimization method and system considering wind and light uncertainty, and the method comprises the steps: carrying out the modeling of a data center energy supply system, generating a wind and light output sample based on Monte Carlo simulation, enabling the generated random sample to meet wind power and photovoltaic output complementation through Corisky decomposition; introducing a first-order autoregression model, generating random scenes in combination with multivariate normal distribution, and obtaining a wind and light output curve in each scene; a two-stage stochastic optimization model containing computing power scheduling and multi-energy coordination is constructed by taking the minimum expected operation cost of a system as a target, the model is solved, and an optimal time sequence operation strategy of a computing power task is decided. According to the two-stage random optimization, the adjustment cost caused by prediction errors is covered with low risk premium, and effective support is provided for reliable operation of the data center under new energy output fluctuation and computing power load time sequence mismatch.

The invention relates to a stochastic simulation method considering dimensional deviation and concrete material variability, and belongs to the field of concrete structure analysis. According to the method, accurate simulation of the mechanical properties of the concrete structure under the influence of construction errors and material randomness is realized by determining random factors, establishing a random field matrix, carrying out sample transformation, establishing random models in batches, solving in batches and carrying out post-processing. A random simulation result is used as an output parameter, a random sample is used as an input parameter, and a PRRM-EPR algorithm is adopted to establish a machine learning prediction model. Compared with a traditional method, the method has the advantages that the generation quality of the random sample is improved, the calculation cost is reduced, the randomness in actual engineering can be better reflected, and the established prediction model can directly predict the mechanical response (displacement, load and the like) of the component through actually measured random parameters. The method is suitable for design, evaluation and optimization of a concrete structure, and particularly has a remarkable effect on improvement of structural safety.

This invention discloses a method and system for probabilistic analysis of seismic displacement of heterogeneous slopes based on finite difference, which evaluates the seismic performance of important engineering slopes by generating slope sliding displacement hazard curves. The method and system first determine the statistical quantities of slope soil parameters and select seismic wave records that match the slope site conditions; then, it generates random field samples of soil parameters and iteratively executes stochastic finite difference numerical simulations to calculate the cumulative slope displacement; next, it constructs a slope displacement prediction model using seismic motion parameters and displacement data, and optimizes the seismic motion parameters based on the model's standard deviation; finally, it develops a slope seismic sliding displacement hazard curve within the framework of probabilistic seismic hazard analysis, estimating the displacement value corresponding to the exceedance probability in the target year. This invention can effectively characterize the spatial variability of soil parameters and the uncertainty of seismic loads, and capture the stress-deformation mechanism of slopes under seismic loading, providing an effective approach for the seismic design of highly important engineering slopes.

The invention discloses a comprehensive energy random optimization scheduling method for a multi-energy scene generation park. The method comprises the following steps: constructing a multi-energy coupling condition feature database based on historical source load multi-energy prediction data; inputting day-ahead prediction information in the multi-energy coupling condition feature database, and constructing an improved condition generative adversarial network model; generating a day-ahead random scene set with a multi-energy coupling characteristic based on the adversarial network model; clustering processing is carried out on the day-ahead random scene set to obtain a typical scene set representing multi-energy uncertainty characteristics and probability distribution of the typical scene set; building a park comprehensive energy system structure based on the energy hub model, and building a day-ahead multi-energy scene random optimization scheduling model in combination with the typical scene set and the probability distribution thereof; and solving the stochastic optimization scheduling model by adopting an optimization algorithm to obtain a robust optimization scheduling scheme of the system equipment. Through the method, a multi-energy coupling scene is generated, scheduling resource waste caused by prediction uncertainty is reduced, and robustness of system optimization scheduling is improved.

This invention relates to a method for establishing a time-varying stochastic constitutive model of rusted cold-formed steel based on random fields. Based on corrosion degree parameters and a two-dimensional power spectral density equation, this invention can not only characterize the degree of uniform corrosion but also evaluate the degree of non-uniform corrosion. Furthermore, it can generate a large number of rusted surface morphologies of cold-formed steel that satisfy a certain statistical distribution through computer simulation. Based on the generated rusted surface morphology data, the mechanical properties of the cold-formed steel are numerically calculated using a finite element model. This method imports the large-scale generated rusted surface data into the finite element model of the cold-formed steel through a program, taking into account the discreteness and uncertainty caused by corrosion. The finite element numerical calculation results are highly consistent with the experimental results. Furthermore, this invention proposes a time-varying stochastic constitutive model of rusted cold-formed steel based on random fields, solving the problem that the constitutive model of rusted cold-formed steel is still based on deterministic experiments, and realizing the prediction of the time-varying stochastic constitutive model of cold-formed steel.

This invention relates to the field of geotechnical engineering physical simulation testing technology, and provides a simulation test method and system for tailings dam slope instability, comprising: acquiring a set of statistical characteristic parameters of tailings sediment in a stochastic field model, and constructing a quantitative model of tailings dam spatial variation including a lens body by using a stepwise decomposition method; building a physical test model on a bottom friction test bench based on the quantitative model of tailings dam spatial variation; applying simulated gravity load to the physical test model by starting the bottom friction test bench, and acquiring image data in real time during the test using a high-speed camera; processing the image data based on digital image correlation to obtain displacement and strain fields, and determining the instability catastrophic data of the tailings dam slope under the action of the lens body based on the evolution of the displacement and strain fields.

This application proposes an efficient simulation method and apparatus for non-Gaussian three-dimensional random fields. The method involves setting target random field parameters, constructing point-by-point edge transformations, establishing latent Gaussian correlation inverse mapping, Nyström-KLE discretization of Sobol sampling, projection of the original spectral samples, orthogonal correction, and node block assembly, ultimately outputting the target non-Gaussian random field. This compensates for the complementary fluctuations missed after low-rank truncation, simultaneously considering both the representation of the dominant structure and the recovery of fine-scale fluctuations. By explicitly subtracting the repetitive components of the original spectral samples in the preserved subspace, the spectral residuals only act outside the preserved subspace, thus avoiding disturbance to the preserved mode coefficients, enhancing the mathematical consistency of the hybrid assembly process, and maintaining good consistency between the preset edge distribution and the spatially correlated structure.

The invention discloses a hybrid energy storage capacity configuration and operation joint decision-making method for multiple power markets, which is used for realizing collaborative optimization of a hybrid energy storage system in planning and operation aspects. According to the method, a double-layer stochastic programming model is constructed, wherein an upper-layer programming model takes annual net income maximization of the hybrid energy storage system as a target, and the rated power and capacity configuration of an energy type battery and a power type battery are optimized; and the lower-layer operation model optimizes a day-ahead market declaration strategy and a real-time response power strategy of the hybrid energy storage system under a day-ahead-real-time two-stage random scene set by taking daily operation expected net income maximization as a target. And the upper-layer planning model and the lower-layer operation model carry out iterative interaction through capacity parameters and operation feasibility information, so that combined optimization of capacity configuration and a multi-market operation strategy is realized. The technical problem that in the prior art, it is difficult for a hybrid energy storage system to achieve planning operation collaborative optimization and reasonable distribution of internal power under multiple uncertainties is solved.

The invention belongs to the technical field of electric power engineering. According to the robust stochastic optimization scheduling method and system for the virtual power plant, source load historical data of a zero-carbon multi-energy system are obtained, the source load historical data are input into a deep learning source load prediction model, and a source load prediction result is obtained; capturing the randomness of the source load prediction result through an auto-encoder, generating a plurality of random scenes, and performing reduction processing on the plurality of random scenes to obtain a target scene set; substituting the target scene set into the target function and the constraint condition, and executing robust random optimization scheduling calculation to obtain equipment operation control parameters of the virtual power plant; outputting a day-ahead scheduling scheme of the virtual power plant according to the equipment operation control parameters; the source load prediction accuracy, the scene reliability, the scheduling scheme multi-target balance capability and the system operation stability are improved, and the problems of unreasonable scheduling scheme and high operation risk caused by prediction deviation, scene redundancy or insufficient constraints are avoided.

The invention discloses a non-stationary corrosion pipeline random field generation method for trend texture reconstruction, and relates to the field of steel pipe corrosion morphology modeling. The method comprises the following steps: firstly, carrying out three-dimensional scanning and filtering processing on a real corrosion pipeline to obtain a pure corrosion height field, and decoupling the pure corrosion height field into a macroscopic non-stationary trend field and a microscopic stationary texture field; thirdly, constructing a mapping relation database of microscopic roughness parameters and power spectral density, and generating a simulated stationary texture field by using a conditional diffusion model; meanwhile, parametric modeling is carried out on the target trend field by adopting an analytic function; determining a modulation coefficient, and performing amplitude modulation and space reconstruction on the target trend field and the simulated stationary texture field; and finally, reversely deriving an optimal trend control parameter based on a constraint genetic algorithm, and generating a final non-stationary corrosion pipeline random field. According to the method, through decoupling and reconstruction of the trend and the texture, the problem that corrosion non-stability cannot be simulated in the prior art is solved, and an accurate data basis is provided for integrity evaluation of the pipeline structure.

The invention belongs to the technical field of freezing temperature fields, and particularly discloses a shield tail brush replacement freezing temperature field stochastic analysis and design method considering soil parameter spatial variability, and the method comprises the following steps: obtaining engineering parameter and soil thermal parameter basic data; establishing a deterministic freezing temperature field model and performing benchmark analysis; constructing a soil thermal parameter 3D random field; performing 3D random freezing temperature field analysis based on Monte Carlo simulation; evaluating the thickness statistical characteristics and reliability of the frozen wall; determining a design expansion coefficient; the invention discloses freezing scheme optimization design considering random characteristics. According to the method, the three-dimensional random thermal value model fusing the random field theory and Monte Carlo simulation is established, the soil variability is systematically considered, the safety and economy of shield tunneling machine tail brush replacement operation are improved, and the limitation of traditional 2D cold region frozen soil stochastic analysis is broken through.

The present application belongs to the field of water and electricity dispatching operation, and relates to a cascade hydropower random intelligent dispatching method driven by multi-agent reinforcement learning. The steps comprise: step one: generating multiple groups of runoff random scenes by using Monte Carlo forward simulation; step two: constructing a two-stage random optimization model with the maximization of power generation benefit as the target based on the scene method; step three: modeling the single-agent reinforcement learning in the spatial dimension of the cascade hydropower; step four: solving the two time period sub-problems decomposed in the time dimension by the single-agent of the cascade hydropower; step five: proposing a multi-agent reinforcement learning step-by-step optimization algorithm MARL-POA, and deploying the reinforcement learning agents with different strategies to cooperatively optimize the dispatching; step six: obtaining the rolling time domain cascade hydropower dispatching scheme as the available information is updated. The present application performs the time-space dimension reduction of the cascade hydropower optimization, effectively reduces the calculation scale, and can quickly and dynamically formulate a reliable cascade hydropower operation plan under the runoff forecast uncertainty.

The application discloses a kind of considering the random programming method of multi-port flexible charging system of electric vehicle charging power dynamic limit.This method includes: collecting power distribution network data, identifying power supply subarea functional attribute and peak charging demand, combined with Voronoi diagram division service range;Build multi-port flexible charging system operation model and charging regulation model considering SOC dynamic limit;Monte Carlo sampling is used to generate joint random scenario set, and representative random scenario and weight are extracted by hierarchical PCA and Wasserstein distance forward selection;With the minimum annual total social cost as the goal, a random programming model containing power distribution network power flow, charging regulation and system operation constraints is established, which is converted into MISOCP by second-order cone relaxation and solved, to obtain power module capacity and charging port number configuration scheme.The application can significantly improve the planning economy and system adaptability.

This invention relates to a method and system for simulating the spatial variability of concrete based on stochastic discrete element method (DEM). The method includes: modeling the concrete material using the DEM based on the Monte Carlo approach, assigning a corresponding particle contact model according to the material's mechanical characteristics; using a random seed number to characterize the random distribution of aggregates; simulating the spatial distribution characteristics of strength parameters between particles using Monte Carlo simulation based on the random aggregates; and introducing a random field into the spatial variation of strength parameters, changing the internal strength spatial distribution of the material through different random parameters. Compared with existing technologies, the numerical simulation method of this invention, based on the DEM and employing a coupling of direct and indirect methods, fully considers the random distribution of aggregates and the spatial variation of strength parameters in concrete materials, and can better reproduce the mechanical variability characteristics of concrete materials.

The invention belongs to the field of hydropower dispatching operation, and relates to a cascade hydropower random intelligent dispatching method driven by multi-agent reinforcement learning. The method comprises the following steps of: 1, generating multiple groups of runoff random scenes by adopting Monte Carlo forward simulation; 2, constructing a two-stage stochastic optimization model taking power generation benefit maximization as a target based on a scene method; step 3, cascade hydroelectric space dimension single agent reinforcement learning modeling is carried out; 4, solving two time period sub-problems decomposed in the time dimension by the cascade hydroelectric single agent; 5, proposing a cascade hydropower multi-agent reinforcement learning step-by-step optimization algorithm MARL-POA, and deploying reinforcement learning agent collaborative optimization scheduling with different strategies; and step 6, obtaining a rolling time domain cascade hydropower dispatching scheme along with the updating of the available information. According to the method, time-space dimension reduction of cascade hydropower optimization is carried out, the calculation scale is effectively reduced, and a reliable cascade hydropower operation plan can be rapidly and dynamically formulated under the condition that runoff forecast is uncertain.