43results about How to "Realize multi-objective optimization" patented technology

The invention relates to a method for carrying out face three-dimensional reconstruction at any viewing angle on the basis of a self-adaptive deformable model. The method includes the steps of (1) obtaining face image data and screening a face image with high definition as original data, (2) positioning feature points, (3) coarsely estimating the angle of a face according to the positioning result of the feature points, (4) building a face three-dimensional deformable model, adjusting the feature points of the face to be at the same dimension as the face three-dimensional deformable model through translation and scaling and extracting coordinate information of the points corresponding to the feature points of the face to form a sparse face three-dimensional deformable model, (5) iterating face three-dimensional reconstruction by means of the particle swarm optimization algorithm according to the coarsely estimation value of the angle of the face and the sparse face three-dimensional deformable model to obtain a face three-dimensional geometric model, (6)mapping input face texture information in a two-dimensional image to the face three-dimensional geometric model in a texture pasting method after the face three-dimensional geometric model is obtained, so that a complete face three-dimensional model is obtained. The method can be widely used in the field of identity identification.

The invention discloses a micro-grid capacity programming method meeting a coupling type electric cold and heat demand, comprising: establishing a micro-grid capacity energy supply device capacity programming model meeting a coupling type electric cold and heat demand, and a micro-grid capacity energy supply device optimum capacity programming calculation method based on the multi-target particle swarm technology. According to the mode, the method solves the problem that a current programmed micro-grid needs to simultaneously provide power supply and cool supply (or heat supply) for users, and the problem of capacity optimum allocation among different energy supply devices when a coupling relation exists between electric cold supply and electric heat supply. The model realizes multi-target optimization under the condition of considering engineering, economy and security constraints of micro-grid energy supply devices.

The invention relates to an adaptive vehicle following algorithm based on improved model prediction control. The algorithm comprises the steps that 1, a vehicle following model is established, whereina controller of an adaptive cruise control system is divided into an upper-layer controller and a lower-layer controller for control, the upper-layer controller calculates expected acceleration according to received information of a relative distance and a relative velocity and transmits the acceleration to the lower-layer controller, and the lower-layer controller controls throttle opening and brake pressure according to a vehicle inverse longitudinal dynamic model through the acceleration; and 2, an algorithm based on model prediction control is established, wherein an estimator is constructed, a least square method is adopted to fit out a most approximate straight line by use of values of previous moments, values of future moments are estimated, and finally a model prediction algorithmis utilized to calculate optimal expected acceleration. Under an existing model prediction framework, front vehicle acceleration information is collected, the least square method is utilized to fit out the front vehicle acceleration change law, prediction is made, a disturbed value of acceleration is provided for model prediction control, and therefore the effect of improving an optimal solutionis achieved.

The invention discloses a micro-grid energy management distribution type multi-target cooperation optimization algorithm based on a potential game. The algorithm comprises the following steps of 1) micro-grid component unit modeling which includes a determination decision main body, a decision variable, a component constraint and multi-target modeling; 2) potential game modeling, wherein an incomefunction and a potential function are vector functions; and 3) based on the distributed game solving of the multi-target optimization algorithm, mainly using a multi-target evolutionary algorithm tosolve the pareto optimal solution set of the income function, and through distributed iteration, solving nash equilibrium. In the invention, a distributed multiagent decision optimization mode is adopted, a potential game theory possessing a distributed characteristic is combined, the expansion of a microgrid is easy to realize, local management and local decision making are realized too, system reliability and flexibility are increased, through a game mode, the distributed multi-target optimization is realized, the competition and cooperate relation of the microgrid and an individual is achieved, and individual interests are guaranteed and the overall benefits of the microgrid are maximized.

The invention discloses a method for optimally selecting thighbone prostheses based on material performance multi-objective optimization, and relates to the field of thighbone prosthesis design. The method comprises the step of utilizing the finite element analysis method for conducting the multi-objective optimization on the thighbone prostheses so as to achieve optimal selection of individualized thighbone prosthesis. The method is characterized by comprising the steps of extracting a thighbone internal contour line and a thighbone external contour line, structurally designing a thighbone prosthesis handle, structurally designing a ball head, assembling a thighbone model and a prosthesis model, selecting thighbone prosthesis materials, establishing a finite element model, setting determined material parameters, constructing a material matching mode and loads, and optimizing evaluation indexes and multiple objectives. The material matching scheme of the ball head and the prosthesis handle is the optimal scheme obtained on the premise that the thighbone prosthesis bearing capacity, the thighbone prosthesis service life, the thighbone prosthesis stress shielding and thighbone prosthesis deformation are optimized under the individual thighbone mechanical environment, the problem of optimally selecting the artificial prostheses is solved by initially applying the fuzzy matter-element theory, and due to the introduction of the method, the multi-objective optimization design of the thighbone prostheses is achieved.

The invention discloses an optimization method for a QoS (Quality of Servie)-based trusted Web service composition. The method comprises the following steps: firstly, evaluating the credibility of the services by analyzing a historical behavior of the services, so as to remove bad services that do not meet the requirement, and obtain the QoS-based trusted Web service as an alternative service for the global optimization of the subsequent service composition; and then using an improved multi-objective Cuckoo search algorithm to realize the multi-objective optimization of the Web service composition, so as to improve the optimization efficiency of the service composition, and finally obtain a composite service that meets the user need and ensures the credibility of the service.

The invention discloses a structure parameter optimization method and device of a wireless charging magnetic coupling device. The method comprises the following steps: building a magnetic circuit model according to the magnetic coupling device to be optimized for obtaining an incidence relation between electromagnetic induction parameters of the magnetic coupling device and preset structure parameters of the magnetic coupling device; acquiring constraint conditions set for the electromagnetic induction parameters and the structure parameters; and taking the incidence relation as a target function, and screening out structure parameter values meeting the constraint conditions through iteration. According to the magnetic coupling device, the defects that a traditional magnetic coupling device is weak in coupling capacity and low in transmission power can be effectively overcome; the electric energy transmission effect of the magnetic coupling device can be effectively improved; the complex design process of optimizing the size of the magnetic coupling device can be simplified; the design efficiency is improved; and the size and the weight of the magnetic coupling device can be effectively reduced.

Provided is a niche sorting particle swarm algorithm based dynamic characteristic optimization method for an electromagnetic mechanism, and belongs to the technical field of dynamic characteristic optimization of electromagnetic mechanisms. The invention aims to solve the problem that at present, only single-objective optimization can be carried out in an optimization algorithm that is used to solve a dynamic characteristic of the electromagnetic mechanism. The optimization method comprises the steps of: firstly, determining dynamic characteristic optimization parameters and an optimization objective function; secondly, determining upper and lower limits of each dynamic characteristic optimization parameter and an additional constraint index that relates to the dynamic characteristic optimization parameter; thirdly, obtaining initialized data of each dynamic characteristic optimization parameter; fourthly, performing calculation to obtain a corresponding optimization objective function value; fifthly, selecting an individual optimal position of a particle parameter and a global optimal position of a whole population, so as to obtain Pareto solution set distribution of the optimization objective function; and sixthly, carrying out measurement or biased selection on the optimization objective function value in the Pareto solution set distribution obtained in the fifth step. The optimization method is used for dynamic characteristic optimization of the electromagnetic mechanism.

The invention belongs to the technical field of building indoor environment design, and discloses a passive energy-saving reverse design system and design method for a high-rise housing, and the method comprises the steps: determining a design parameter variable and a design target according to a building design type; analyzing the influence of different factors on the thermal environment and theenergy consumption condition of the building by utilizing the test data; analyzing influence rules of passive factors on building thermal environment and energy consumption under different combinationconditions by adopting a computer simulation method in combination with the building information model; determining influence characteristics and value ranges of different passive factors on buildingenergy consumption; establishing a rapid prediction model to analyze the influence relationship of the high-sensitivity parameters on the building thermal environment and energy consumption; obtaining a design scheme solution set meeting a target requirement based on the rapid prediction model; and establishing a multi-objective optimization model, carrying out data fitting, and outputting a passive energy-saving strategy of the high-rise residence. According to the method, the reverse design convergence is ensured, and meanwhile, the reverse design calculated amount is reduced by 40%.

The invention provides a jet pump multi-objective optimization method based on a neural network model and an NSGA-II genetic algorithm. The method mainly comprises the steps that jet pump design parameters, optimization objectives and constraint conditions are determined, and sample point design parameters are obtained based on a sampling method; an optimization target value corresponding to the sample point design parameter through CFD software simulation is acquired. the sample point data is used to construct a neural network model of jet pump design parameters and an optimization target, and prediction precision is verified. based on the neural network model, an NSGA-II genetic algorithm is adopted to obtain a final optimization result. the CFD method is combined with the neural network model and the genetic algorithm, the problem that complex optimization design is difficult to solve due to multiple parameters and multiple disciplines is solved, the calculation difficulty is reduced, the problems that an optimization design method based on CFD simulation or experiments is high in cost and long in consumed time in the past are solved, multi-target optimization of the jet pump is achieved. The special design requirement for the lift ratio in actual engineering is met, and the hydraulic performance of the jet pump is effectively improved.

The invention discloses an RSM-based motor multi-objective optimization design method. The method comprises the following steps of S1, establishing an RSM fitting model; and s2, carrying out orthogonal optimization. According to the RSM-based motor orthogonal optimization method, the relation between system response and input variables is obtained by analyzing a small amount of test data, a systemmathematical model is established, on the basis, an orthogonal design method is adopted to design a test scheme, the established mathematical model is applied to predict a test result, and orthogonalanalysis and optimization are carried out on the prediction result.

The invention provides a temperature control method of a power battery, an AMPC controller, a thermal management system and a medium. The thermal cycle topological structure of the power battery comprises a battery pack and a temperature control device. The temperature control method comprises the following steps of: establishing a temperature control prediction model of the power battery according to a thermal cycle topological structure of the power battery; respectively acquiring the predicted temperature of the battery pack and the energy consumption information of the temperature control device by using the temperature control prediction model and a plurality of groups of preset temperature control parameter information; according to a preset optimization performance index and a target function, selecting one group of temperature control parameter information from the plurality of groups of temperature control parameter information as control input of the temperature control device; and correcting the temperature control prediction model in real time according to the obtained state parameters of the battery pack. According to the temperature control method of the power battery, the AMPC controller, the thermal management system and the medium provided by the invention, the temperature of the battery pack of the power battery can be controlled to be within/close to an appropriate range under the condition of not increasing any hardware cost, and the energy consumption of an actuator can be reduced.

The invention discloses a method for controlling the voltage of a photovoltaic power station based on neural networks. The control method comprises the following specific steps that step 1, a bus voltage active power neural network and a bus voltage reactive power neural network respectively output corresponding sensitivity coefficients of bus voltage active power and bus voltage reactive power toa model prediction controller; and step 2, the model prediction controller constructs a mathematical model of bus voltage and node injection active power and reactive power via the sensitivity coefficients of the bus voltage active power and the bus voltage reactive power issued by an upper-layer control architecture. According to the method for controlling the voltage of the photovoltaic power station based on the neural networks, the overvoltage and undervoltage problems of the bus of the photovoltaic power station can be solved, the practicability and expandability of the method based on sensitivity analysis are effectively improved, and the method has important significance for safe operation of a power distribution network and can promote energy internet construction, improve power service quality and optimize utilization of various resources.

The invention relates to a method for selecting the optimal grid connection position of a distributed power source, which is characterized in that the method comprises the following contents: establishing a distribution network model containing the distributed generation, establishing a model for optimizing the configuration of grid-connected distributed generation based on a pseudo physical algorithm, establishing a multi-objective optimization process, and the like. Through calculation and analysis, the optimal installation position of distributed generation in distribution network and the optimal analysis of its admittance capacity after fixing the installation position of distributed generation are obtained,the methodhas strong robustness, adaptability and high practical application value.

The invention provides a circulating current suppression direct prediction control method and system for a parallel converter cluster. The method comprises the following steps of: obtaining a three-phase current value of a converter cluster at the current moment, and carrying out coordinate transformation and delay compensation; predicting a current value at a subsequent moment according to the compensated three-phase current value, calculating cost functions according to the predicted value, comparing all the cost functions, and selecting a voltage vector minimizing the values of the cost functions as an optimal voltage vector; for the compensated three-phase current value, calculating a zero-sequence circulating current, and selecting a zero voltage vector according to the direction of the zero-sequence circulating current at the current moment k; and according to the obtained optimal voltage vector and the zero voltage vector, constructing a synthetic voltage vector, and obtaining a switching signal sent to the converters to eliminate the circulating current. On the premise of accurately controlling the output current, the zero-sequence circulating current between the parallel converters can be effectively eliminated.

The invention belongs to the technical field of solar hybrid power generation, and particularly provides a photovoltaic photo-thermal complementary power generation system optimal configuration method, which comprises the following steps: building a subsystem model, carding boundary conditions of the subsystem model, and setting operation limiting conditions of the subsystem model on a time sequence; setting an optimization target of the subsystem; a swarm intelligence algorithm is applied to encode variables of the subsystem, simulation is performed through a subsystem model after encoding, and a subsystem fitness function is obtained based on a simulation result; performing iterative optimization on the fitness function of the subsystem according to the target of the subsystem; and judging whether the iterative optimization is terminated or not, and configuring the subsystem according to the optimized subsystem fitness function. The problem that an existing photo-thermal power generation system cannot achieve optimal configuration of the system is solved, reasonable technical configuration of the photovoltaic photo-thermal system can be achieved, and parameter configuration of a photo-thermal power station can meet the requirement for peak regulation of the system.

The invention discloses a household equipment energy consumption multi-objective optimization method and system based on an IBAS algorithm, and the method comprises the steps: dividing typical household equipment into a schedulable interruptible load, a schedulable non-interruptible load and a non-schedulable load based on a controllable household load classification method, and establishing user demand side load scheduling models respectively; according to the on-off state of the equipment in each time period obtained by the user demand side load scheduling model and corresponding constraint conditions, establishing a multi-target optimization scheduling model of the household energy management system by taking the lowest household user power consumption cost, the minimum load peak-to-average ratio and the maximum user comfort as targets; converting a multi-objective optimization scheduling model of the household energy management system into a single-objective function to serve as an optimization function of the system by adopting a weighted combination and multiplication and division method, and solving the optimization function of the system by adopting an improved IBAS algorithm to obtain equipment power utilization arrangement of the next day of a family for day-ahead optimization scheduling. According to the method, the comfort level of the user is improved, and multi-objective optimization of the household energy consumption equipment is realized.

The invention discloses a household energy multi-target optimization method based on GA-BFO. The method is characterized by according to a load characteristic of household electrical equipment and a user demand, classifying the household electrical equipment, and establishing a control model and an optimization constraint for schedulable equipment; establishing three optimized target functions ofminimum power cost, minimum PAR and maximum user comfort; establishing a target function expression of load transfer and solving the target function expression through using a GA-BFO algorithm to obtain next day arrangement of household schedulable equipment; and in a scheduling day, when a system receives a rescheduling request of a user, recording information of the equipment which needs to be rescheduled, establishing the target function expression, converting a real-time scheduling problem into a knapsack problem, and using a dynamic equation to solve, obtaining real-time scheduling arrangement of the equipment which needs to be rescheduled so as to complete multi-target optimization. The method is used to coordinate and manage the schedulable electrical equipment in a family and an electricity utilization mode of an electric vehicle, which can effectively reduce household electricity cost, decrease PAR and improve user comfort.

The invention discloses a data center virtual machine placement method and system, a medium and equipment, N virtual machines and M physical machines are given, p sequences po are initialized at the same time, and each sequence represents a sequential request sequence of the N virtual machines and serves as an original population; a physical machine in which the virtual machine should be placed is selected for each sequence pi in the original population by adopting an optimal adaptation method, a temperature rise index is calculated after the virtual machine is placed through an evaluation function, and the physical machine with the minimum temperature rise index in the current sequence pi serves as a current target physical machine; an initial population of the genetic algorithm is obtained through p times of circulation; crossover and mutation operation is performed on each sequence in the obtained initial population, the fitness of each sequence is calculated through a fitness function, and the population of the next iteration is selected by using a roulette algorithm, after t rounds of iteration, a physical machine sequence with the minimum fitness function value is selected to serve as a sequence with the minimum energy consumption and the most average temperature distribution, and virtual machine placement is completed; energy saving and hot spot avoiding targets can be achieved.