30 results about "Kriging algorithm" patented technology

The invention discloses a deterministic method for qualitatively identifying the top and bottom interfaces of hydrate in the sea, which includes the following steps: A, getting multiple sensitive attributes for identifying hydrate through seismic reflection characteristic research, AVO, elastic wave impedance inversion and other research methods; B, calculating the weight coefficients of the attributes by analyzing the relationship between the sensitive attributes and hydrate characteristics, and getting attributes used to identify the hydrate characteristics through information fusion analysis; C, studying the spatial correlation between the weighted attributes and sonic skew log variables, and building a model capable of reflecting the hydrate characteristics through a collaborative Kriging algorithm based on the information of the weighted attributes; and D, on the basis of collaborative Kriging modeling, carrying out global optimization broadband constrained inversion to qualitatively identify the top and bottom interfaces of hydrate. The limitation brought by a single technology and the multi-solution interference are reduced. A favorable target of hydrate can be found more accurately, and the top and bottom interfaces of hydrate can be identified.

The invention discloses a method and system for optimizing energy consumption of a pan-park air-conditioning system based on the RSM-Kriging-GA algorithm. The method includes: reading the current course arrangement, setting the start and stop of the air conditioner according to the course arrangement; automatically selecting according to the date of the day According to the corresponding seasonal mode, obtain the historical data under the corresponding seasonal mode; use the RSM cross model to fit the relationship between the number of courses, ambient temperature, classroom volume and real-time cooling equivalent in the historical data; obtain the current environmental parameters in real time, according to the RSM Calculate the real-time cooling equivalent required by the cross model; construct the constraint relationship, and use the kriging algorithm to establish a mathematical model between environmental parameters and energy consumption E; solve the optimal parameter solution based on the GA algorithm, and set the temperature and wind speed according to the optimal parameters , Refrigeration time to optimize energy consumption and automatic regulation. The present invention fits historical big data based on the RSM model, constructs a constraint model, and establishes an energy consumption optimization model of a pan-park air-conditioning system based on the RSM-Kriging-GA algorithm, thereby optimizing energy consumption of the air-conditioning system and saving resources.

A universal mobile sensing and forecasting method for regional environmental information, which uses mobile sensing nodes to collect data and transmits it to the cloud server; Then use the improved stochastic L-BFGS algorithm to improve the Kriging algorithm, and use this algorithm to carry out universal regional environmental information spatial prediction, and obtain the spatial prediction results of environmental information; use the shared The relationship between the time and space of the environmental data collected by the bicycle is used for spatio-temporal prediction and analysis, and the spatio-temporal prediction result of the target area is obtained. Finally, the final regional global environment prediction result map is synthesized based on the different prediction results of time and space.

The invention provides a method for determining a well location in a high water-cut oil field based on an evolutionary algorithm, comprising the steps of: establishing a reservoir model of a predetermined oil area, dividing the reservoir model into grids; Combined to obtain the remaining oil potential enrichment area; calculate the remaining oil potential enrichment area in the remaining oil potential enrichment area, which represents the distribution of remaining oil and the well location constraint value of the injected water sweep volume, and obtain the well location advantage area; use the evolutionary algorithm to iterate the oil wells in the advantage area Calculate the fitness function, use the kriging algorithm to evaluate the fitness function, the evolutionary algorithm is adjusted according to the fitness function, the kriging algorithm considers multi-dimensional space, nugget effect and local weight; iteratively calculates according to the evolutionary algorithm, and obtains the maximum fitness degree function value to determine the well location. The invention uses the improved evolutionary algorithm and Kriging algorithm as optimization tools, constructs a dynamic potential index to narrow the search space of the evolutionary algorithm, and greatly increases the convergence accuracy and speed of optimization calculation.

The invention provides a multi-working-condition structure optimization method of a weapon station on the basis of a Kriging algorithm. The method comprises the following steps: 1: giving the range of a support frame elasticity modulus, the range of cartridge box quality and the range of buffer rigidity; 2: adopting Latin hypercube experiment design to select and obtain N groups of sample points; 3: calculating gun muzzle vibration comprehensive parameters; 4: combining the N groups of sample points with N gun muzzle vibration comprehensive parameters to form an initial training sample point set; 5: using a genetic algorithm to carry out optimization on a Kriging surrogate model to find an optimal point and a maximum expectation improvement point; 6: on the basis of a dual Kriging model sequence iterative optimization algorithm, carrying out optimization on a structure parameter optimization problem until convergence is realized; and 7: obtaining the structure optimization parameters of different angles of fire, and substituting the optimal structure parameter of each angle of fire into weapon station dynamics models under other working conditions, and looking for the structure parameter with a maximum average optimization range as a final structure optimization parameter.

The invention discloses an increment ionosphere refraction error correction method which is applied to a GNSS single-frequency user. The method comprises a monitoring station utilizes ionosphere model parameters broadcasted by a GNSS system itself to calculate ionosphere vertical TEC1, and meanwhile, utilizes double-frequency observation data to calculate ionosphere vertical TEC2, and difference value [delta] TEC between the TEC1 and the TEC2 is obtained; a plurality of monitoring stations transmit the [delta] TECs of themselves to a master control station through a communication link; the master control station receives each [delta] TEC, establishes an ionosphere increment variation model, obtains ionosphere increment variation model broadcast parameters a, b and c in a region, and transmits and broadcasts the a, b and c to the user through a broadcast link; a user receiver receives a navigation signal and the a, b and c, utilizes the ionosphere model parameters broadcasted by the GNSS system itself to calculate TEC <GNSS>, utilizes the a, b and c to estimate optimum estimation value (shown in the specification ) of the ionosphere increment through a common Kriging algorithm, and adds the TEC <GNSS> with the optimum estimation value (shown in the specification ) to obtain TEC ; and the TEC is utilized to calculate ionosphere addition time delay, so that high-precision ionosphere refraction error correction for the GNSS single-frequency user is realized.

The invention provides an overhead weapon station structure optimizing method based on an improved Kriging algorithm. The method includes the following steps of firstly, presetting the ranges of the elasticity modulus of a supporting frame, the mass of an ammunition can and the rigidity of a buffer; secondly, designing, selecting and obtaining N sets of sample points through a Latin hypercube test; thirdly, calculating vibration comprehensive parameters of gun muzzles; fourthly, establishing an initial training sample point set by combining the N sets of sample points with the vibration comprehensive parameters of the N gun muzzles; fifthly, optimizing a Kriging agent model through a genetic algorithm, and finding the optimal point and the maximum expectation improvement point; sixthly, optimizing the structural parameter optimizing problem on the basis of a duplex Kriging model sequence iterative optimizing algorithm till the algorithmconverges. By means of the method, through the optimization design of the elasticity modulus of the supporting frame, the mass of the ammunition can and the rigidity of the buffer, the disturbance quantity of the gun muzzles is made the minimum.

The invention provides a weapon station muzzle disturbance optimization method based on the Kriging algorithm. The method comprises the steps of: firstly, setting the ranges of the buffer damping, the barrel elasticity modulus and the barrel Poisson ratio of an overhead weapon station; secondly, selecting N sample points by using the Latin Hypercube experiment design; thirdly, calculating a muzzle disturbance comprehensive function; fourthly, combining the N groups of sample points and N muzzle vibration comprehensive parameters to form an initial training sample point set; fifthly, performing optimizing on a Kriging surrogate model by using the genetic algorithm to find out the optimal point and the expectation maximum raising point and obtain optimized structure parameter values. The weapon station muzzle disturbance optimization method based on the Kriging algorithm can minimize the muzzle disturbance quantity through the optimization design of the elasticity modulus of a support frame, the quality of a cartridge box and the rigidity of a buffer.