111results about How to "Achieve global optimization" patented technology

The invention relates to a real-time path planning method of AUV (Autonomous Underwater Vehicle), in particular to a method for carrying out online, real-time local path planning according to an online map in an AUV real-time collision preventation process. The method comprises the steps of: setting the quantity of small populations according to the quantity of path points of the AUV, initializing; carrying out immune selection on each small population to obtain subgroups; carrying out genetic manipulation on one subgroup, carrying out cell cloning on the other subgroup; then clustering through a vaccination and an antibody to form the next generation of small population, judging whether the next generation of small population meets the conditions or not; if yes, selecting optimal individuals of the small populations; and selecting the optimal individuals from the set consisting of all optimal individuals to be used as a planning path. According to the invention, the diversity of the population is maintained by using an antibody clustering principle, the premature convergence of an algorithm is avoided, and the global optimization is facilitated. The established immune genetic algorithm is used for clustering and analyzing generated filial generations by adopting a self-regulating mechanism, and the diversity of the population is ensured.

The invention discloses a method for regional water-source optimized configuration based on an MAEPSO algorithm. The method comprises the steps that basis information data of a water resource system is acquired at first; secondly, a water-resource multi-target configuration model is established; then the algorithm based on MAEPSO is executed, and a Pareto optimal solution set of multi-target configuration of the regional water resource system is solved; and finally, a final solution is selected from the optimal solution set according to certain rules. According to the invention, global optimization is achieved; computation efficiency is increased; and a requirement for selecting a multi-target optimal configuration scheme of the water resource system can be satisfied.

The invention discloses a method for solving low-carbon workshop scheduling based on an improved whale optimization algorithm. The method comprises the following steps: establishing a mathematical model of low-carbon workshop scheduling; setting algorithm parameters of the improved whale optimization algorithm, and generating an initial population; calculating the fitness value of the scheduling solution in the initial population, and keeping the current optimal scheduling solution; converting the current optimal scheduling solution into a whale individual position vector; carrying out whale individual position vector iteration updating by adopting an improved whale algorithm; performing whale individual position vector iteration updating on the updated whale individual position vector byadopting a self-adaptive adjustment search strategy; and when the number of iterations reaches the maximum number of iterations, converting the whale individual position vector into a scheduling solution, and outputting the scheduling solution. The whale algorithm is optimized, and a two-stage conversion mechanism is applied to initialize a machine part and a process part respectively, so that thenumber of iterations is reduced, and the quality and the operation efficiency of a final solution are improved; an improved whale algorithm is adopted, and the convergence speed and efficiency are improved.

Provided are a parameter control method and a parameter control system. The method comprises the steps of the relevance of parameters to be detected is built, and an initial threshold range is set for each parameter to be detected; the value of each parameter to be detected is acquired, and whether the values are within the corresponding initial threshold ranges or not is obtained in a comparing mode; when the values exceed the corresponding initial threshold ranges, the initial threshold ranges of the parameters to be detected or other to-be-detected parameters related to the parameters to be detected are adjusted according to the relevance of the parameters to be detected. The system comprises a setting module, an acquiring module, a comparing module and an optimization module, the setting module is used for building the relevance between the parameters to be detected and respectively sets the initial threshold ranges for the parameters to be detected. The acquiring module acquires the value of each parameter to be detected. The comparing module judges whether the values are within the corresponding initial threshold ranges or not. The optimization module adjusts the initial threshold ranges of the parameters to be detected or other to-be-detected parameters related to the parameters to be detected according to the relevance of the parameters to be detected. The parameter control precision can be improved, and finally production cost is lowered.

The invention relates to the field of natural language processing and particularly relates to an information extraction system based on bidirectional RNN. The system comprises a word segmentation module, a dictionary mapping table, a bidirectional RNN module and an information extraction module, wherein the word segmentation module, the dictionary mapping table, the bidirectional RNN module and the information extraction module are sequentially connected. A classification sequence of information corresponding to an inputted text sequence is predicted through the bidirectional RNN module and a corresponding entity name in the classification sequence is extracted through the information extraction module. When the bidirectional RNN module used by the system predicts the entity name, vector information converted by the to-be-processed text sequence is firstly forwardly and then backwardly inputted to the bidirectional RNN module at a corresponding moment, prediction of the classification result at each moment depends on historical information and also depends on future information, the predicted result is more accurate and reasonable, and the system has good application prospect in the data processing field, and particularly, in the fields of entity name extraction and recognition.

The invention relates to the field of natural language processing, in particular to an information extraction method based on a bi-directional recurrent neural network. The information extraction method applies the technology of the bi-directional recurrent neural network, the basic elements, which include characters, words, punctuations and the like, of a text are subjected to the automatic learning of characteristics, series modeling is carried out through the RNN (Recurrent Neural Network), and the defect that the characteristics need to be manually set in a traditional way is overcome. In addition, the bi-directional communication RNN is used to overcome the problem of information asymmetry in a prediction process of a unidirectional RNN, so that the classification judgment result of a natural language series to be identified depends on both preamble information and postamble information, and therefore, information extraction and judgment accuracy is higher. The method is especially suitable for entity name extraction in big data analysis, and has an important application value in the big data analysis.

The invention belongs to the technical field of internet advertisement, and provides an internet advertisement real-time bidding multi-advertiser multi-factor flow distribution method. The method comprises the steps that an advertisement bidding unit ad group is divided into an offline multi-factor mixed grading function module, an offline maximum cost estimation function module, a grade cumulative distribution estimation function module and an online flow distribution function module. In the offline multi-factor mixed grading function module, offline multi-factor mixed grading is carried out on advertisements; in the offline maximum cost estimation function module, offline maximum cost estimation is carried out on the advertisements; in the grade cumulative distribution estimation function module, grade cumulative distribution estimation is carried out on PV flow participating in biding; in the online flow distribution function module, online flow distribution is carried out on the advertisements. Numerical values obtained through the steps are compared, and a conclusion of whether quotation is carried out or not is obtained. The problems that some advertisements have fewer and fewer putting chances, the budget cannot be spent, it is unfair for advertisers, and DSP overall situation putting earnings are affected are solved.

The invention belongs to the field of intelligent transport study and relates to an adaptive intersection signal control method considering supply side constraint. The problem that consideration on intersection exit ramp transportation supplies is absent in the existing signal control system and study is solved. The method comprises the following steps: 1, detecting and processing intersection traffic information, and acquiring traffic demands of an intersection entrance ramp and traffic supplies of the exit ramp; 2, establishing a time-varying relationship of intersection traffic supplies, and constructing an intersection supply-demand matrix; 3, establishing an intersection signal timing dial optimization model for quintuple descriptions on problem description, initial state, target state, action set and knowledge rule based on intelligent planning; and 4, searching the target state by adopting a decision-making tree method, and generating a corresponding timing plan. Compared with the prior art, the method disclosed by the invention has the advantages that active control of intersections is realized, the intersection downstream exit ramp queue overflow can be avoided, and intersection 'deadlock' is avoided.

The invention provides a resource allocation method, device and equipment, and the method comprises the steps: obtaining a resource application request for a specific service, wherein the resource application request comprises resource information; selecting resources from a mixed resource pool according to the resource information; allocating the resources selected from the mixed resource pool tothe specific service, wherein the mixed resource pool is used for providing resources for at least two services. By means of the technical scheme, mixed deployment of resources is achieved, the purpose of improving the resource utilization rate is achieved, and finally resource sharing is achieved.

The invention provides a bearing fault diagnosis method and system. The bearing fault diagnosis method comprises the following steps: collecting experimental data of a faulty bearing under a working condition, sequentially constructing a training set, and establishing a classification model by using the training set; optimizing the smoothing factor of classification model by a verification set anda test set to obtain an optimized model; finally, performing fault diagnosis on bearing operation data by the optimized model. The classification model has higher classification accuracy and higher training and updating speed.

The invention belongs to the field of power distribution network fault restoration algorithms, and discloses a major network and island synchronization fault restoration algorithm for a power distribution network including DGs. Based on an original power distribution system, virtual branch circuits are additionally arranged between the DGs and transformer substation nodes so that an expansion network can be formed, major network reconstitution, island generation and DG operating mode determination can be synchronously processed by executing a primary genetic algorithm in the following computation process, and thus global optimization can be achieved; then, the system enters in the genetic algorithm iterative process, a load shedding operator is introduced, and the specific shedding mode is decided according to the type and the position of the load to be shed; in addition, after each iteration, an infeasible network structure in a new population further needs to be amended, the network is kept to be a fully-communicated radial network, and the loss is lowered minimally after the power distribution network breaks down.

The invention discloses an integrated static compensator of a distribution transformer, comprising the distribution transformer, a static compensation module and a control platform, wherein each phase winding at a high-pressure side of the distribution transformer is provided with a connecting tap; the connecting tap is connected with an alternating current output side of the static compensation module; and the control platform controls the static compensation module to inject reactive power by the connecting tap so as to realize compensation. The integrated static compensator centralizes to compensate at a key node, thus improving the capacity of the static compensator of existing distribution networks and improving the quality of compensated output waveform.

The invention discloses a method for optimizing the natural gas liquefaction process technological parameters based on the genetic algorithm. According to the method for optimizing the natural gas liquefaction process technological parameters based on the genetic algorithm, on the basis of an original technological process, the genetic algorithm is introduced, a corresponding original genus, a fitness judgment formula, the cross-over rate, the aberration rate and the end condition are established for each set of parameters based on the genetic algorithm in the process of optimization of the process technological parameters, genetic algorithm optimization operation is conducted, and finally the optimal genes are obtained and used as the optimal technological parameters of the natural gas liquefaction process. Under the condition that the inherent characteristics of all devices of a system and other design regulations are met, the energy consumption generated in the natural gas liquefaction process is reduced. Meanwhile, the method for optimizing the natural gas liquefaction process technological parameters based on the genetic algorithm has the advantages of being supermatic in intelligent optimization, capable of improving the optimization efficiency, and high in global superiority and universality. According to the method for optimizing the natural gas liquefaction process technological parameters based on the genetic algorithm, the optimization range covers all the technological parameters of the process, and global optimization is really achieved through interrelation of the parameters. The method for optimizing the natural gas liquefaction process technological parameters based on the genetic algorithm can also be applied to various optimization objects, and therefore the universality of natural gas liquefaction technological parameter optimization is improved.

The invention is suitable for the technical field of computers and provides a dynamic placement optimization method and system of a virtual machine on a server cluster. The method comprises the steps of respectively establishing a server status energy consumption model, a virtual machine migration energy consumption model and a server status switch energy consumption model and establishing a server cluster energy consumption model according to the server status energy consumption model, the virtual machine migration energy consumption model and the server status switch energy consumption model; calculating out average waiting time of user requests according to probability that servers are in each state, amount of current user request servers and amount of the servers in each state; using the minimum value of server cluster energy consumption as an optimization target according to the server cluster energy consumption model under a preset condition to obtain a dynamic placement optimization scheme of the virtual machine. The dynamic placement optimization method disclosed by the invention can model a dynamic change process of a load and a status of the server cluster and utilizes the minimum value of the server cluster energy consumption as the optimization target to obtain the dynamic placement optimization scheme of the virtual machine.

The invention discloses an adaptive echo cancellation method. The method introduces the viewpoint of evolutionary computation in the artificial intelligence, simultaneously employs the cross alternative method and the idea of the function global optimization and combines the idea of time driving, so as to effectively solve the problem that the constant Delta value of the parameter of the conventional NLMS algorithm can not be determined, thus improving the performance of the echo cancellation algorithm NLMS.

The invention discloses an SDN-based space-based information network multi-controller efficacy-losing backup method. The method comprises the following steps of constructing an SDN-based space-based information network multi-controller efficacy-losing backup framework; designing an efficacy-losing detecting mechanism; recovering a controller election algorithm based on faults of a Paxos algorithm.The SDN core technology is applied to a space-based information network, and therefore the structure of network equipment is simplified, a satellite just needs to achieve the simple forwarding and hardware configuring functions, and therefore the defect that satellite nodes are complex in design and high in manufacturing cost is overcome. The SDN-based space-based information network can achieveoverall resource distribution and overall optimization, and the resource utilization rate is increased. By adopting the multi-controller framework, the expansibility of the space-based information network can be improved, and therefore the flexible and efficient network configuration and management are achieved. In addition, centralized management has the overall view of all nodes in the whole network, the state of the satellite nodes can be grasped, and the network destroy resistance is improved.

The invention discloses a photovoltaic control method and device, equipment and a computer readable storage medium. The method comprises steps that the parameter information of a selected distributionnetwork system is inputted; taking each photovoltaic system as the center of a region, the distribution network system is dynamically partitioned according to the information and a time period of nodes in the distribution network system to obtain multiple partition regions; a setting model is established according to the parameter information of the distribution network system and the partition regions obtained by partition; the setting model is calculated to obtain the calculation result. The method is advantaged in that firstly, the distribution network system is dynamically partitioned, secondly, the same type of PV systems in the same partition region are considered as a PV cluster, the same voltage control strategy and the same energy storage system operation strategy are utilized for control, advantages of centralized control are combined with advantages of local control, and thereby not only can the amount of communication information be reduced, but also global optimization between the partitioned regions is achieved, and computational burdens can be reduced.

The invention relates to the technical field of an active power distribution network, and discloses a distributed power supply synergic control method and system, a server and a computer readable storage medium. The method comprises the steps of controlling each region to do power interaction to achieve global optimization according to a global optimization control strategy under long-time scale;acquiring an actual output power value of each region; acquiring a stability estimation index value according to the actual output power value of each region; comparing the stability estimation indexvalue with a preset index value to obtain a comparison value; judging whether synergic control is performed on the output power value of each distributed power supply in each region or not according to a comparison result. After global optimization, the actual output power is estimated by combining the stability estimation index value, an unstable power generated by the distributed power supply ineach region of the active power distribution network is favorably and timely adjusted, so that the integral stability of the power distribution system is ensured.

The invention discloses a double-objective comprehensive optimization design method for LED radiator structure parameters. According to the method, a continuous function relation between a comprehensive optimization target and the structure parameters to be optimized is accurately built through a response surface method, an orthogonal testing method and a genetic algorithm are reasonably combined to use, global optimization can be completed in a solution space of a comprehensive optimization target function of the structure parameters of a radiator in complex appearance, and a Pareto optimal solution achieving comprehensive optimization of double objectives containing improvement of LED cooling performance can be obtained, so that the optimal combination of various structure parameter design values is quickly determined according to actual requirements of the design target of the radiator, cooling efficiency of the radiator is effectively improved, the research and development period is greatly shortened, and product cost in the links of design, manufacturing, application and the like is reduced.

The invention discloses a method for global optimization and distribution of agile satellite imaging tasks and belongs to the field of agile satellite imaging task planning. The method includes analyzing visibility of a satellite for all imaging targets, selecting visible time periods in an inbound pass segment and in which a lateral swing angle that the satellite points to each target start point at an intermediate moment does not exceed a specific range, and sequentially the imaging visible targets in corresponding orbit laps according to the sequence of the visible time periods. The method takes multiple factors such as satellite energies, imaging efficiency and imaging quality from the perspective of satellite application and puts forward the optimization principle that the target imaging time periods are in the inbound pass segment and the lateral swing angle at the intermediate moment of each visible time period is smaller than the specific range, and multiple imaging tasks are distributed to the proper orbit laps to enable the imaging task planning to be more reasonable and meet the satellite application requirements, so that observation results are optimized, imaging quality is guaranteed and energy consumption is reduced.

The invention discloses a comprehensive energy microgrid distributed economic dispatching optimization method and system, and the method comprises the steps: analyzing the working characteristics of a hydrogen energy storage system, and obtaining the energy flow paths of a power energy flow, a natural gas energy flow and a hydrogen energy flow; according to the working characteristics of the hydrogen energy storage system, establishing an economic dispatching optimization model with the purpose of minimizing the operation cost of the integrated energy microgrid; establishing constraint conditions of the economic dispatching optimization model, wherein the constraint conditions comprise hydrogen energy storage system constraint, power system constraint, natural gas system constraint and coupling element constraint; according to the economic dispatching optimization model and the constraint condition, constructing a distributed solving method based on an alternating direction multiplier method, obtaining an optimal solution meeting the convergence condition after the power system economic dispatching optimization sub-problem and the natural gas-hydrogen economic dispatching optimization sub-problem are subjected to alternating iteration solving, and optimizing economic dispatching. Global optimization can be realized under the condition of limited information interaction.

The invention relates to the field of natural language processing, in particular to an enterprise entity name analysis and identification system. The system comprises a bidirectional recurrent neural network module; the system uses an enterprise name labeling training sample stored in an existing enterprise name database to train a bidirectional recurrent neural network; the bidirectional recurrent neural network identifies enterprise names in to-be-processed texts and extracts names which do not belong to existing enterprise names; the system enables a classification and judgment result of to-be-identified natural language sequences to depend on context information and enables the preparation rate of extraction and judgment to be higher by performing characteristic automatic learning on basic elements of the texts, such as characters, words, punctuation marks and the like, and applying the bidirectional RNN (Recurrent Neural Network); and the system discovers new enterprise entity names through existing data characteristics and has important application values in the field of big data analysis, particularly in the field of data analysis taking enterprises as analysis main bodies.

The invention discloses a method and a device for real-time global optimization and error loopback judgment in three-dimensional reconstruction, and the method comprises the following steps: obtainingvisual information and IMU information, and constructing an optimization equation through the secondary statistics of the visual information and the pre-score of the IMU information, so as to carry out the real-time online global optimization of a CPU; coupling a transformation matrix in the global optimization function, and providing a state estimator in the global optimization process, so thatmapping continues to be carried out when rapid movement is carried out and the number of feature points is small; and obtaining an estimator of IMU zero offset, and identifying an error loop and a correct loop according to the estimator of IMU zero offset. According to the method, high-efficiency global optimization of vision and IMU information can be realized on the premise of not losing precision, the problem of accumulative errors caused by inaccurate calibration is solved, the loop-back correctness can be accurately judged, and the method is simple and easy to realize.

The invention provides a global optimization method of a numerical program. The global optimization method of the numerical program includes the steps that a symbol execution technology is adopted toextract algebraic representations for expressing path constraint conditions and numerical calculation processes from a source program; all the algebraic representations are analyzed, and by using different algebraic transformation rules, the algebraic representations are converted to be in algebraic forms with lower numerical calculation errors; finally, all the algebraic representations are converted into corresponding code fragments, and the code fragments are combined to generate a target program. The global optimization method of the numerical program has the advantages that all the processes are automation processes, and no human intervention is needed; a program writer can better concentrate on service logic and does not need to care about detail operation related to numerical analysis. In this way, not only is the development efficiency improved, but also code logic is visual and easy to understand and maintain.

The invention discloses a state of charge of battery planning method for global energy management of hybrid electric vehicles. The method comprises a discrete method of an SOC feasible domain of a vehicle power battery and the planning method of an SOC value, specifically includes a calculation of a corresponding moment of a boundary transition point of the SOC feasible domain, a calculation of aSOC maximum value at each time in the SOC feasible domain, a calculation of the number of discrete points and discrete intervals at each time in the SOC feasible domain, a calculation of the SOC valueunder a corresponding label at each time in the SOC feasible domain and the like. The discrete method achieves reasonable discretization within the feasible domain of a battery SOC based on the limitation of a maximum discrete interval of the SOC and the number of the discrete points of the SOC. The planning method finds an optimal control sequence by planning the SOC value, and achieves global optimality of energy management control. The method both ensures accuracy requirements of battery energy management calculations, and significantly reduces calculation loads and costs, and has high comprehensive performance.

The invention discloses a robot rigidity design method based on a rigidity model and a terminal. The robot rigidity design method comprises the following steps: establishing the robot rigidity model according to robot structure and rigidity design parameters; selecting a reference pose which conforms to the robot rigidity design; constructing a rigidity optimization comprehensive model which comprises rigidity comprehensive evaluation functions, constraint conditions and initial values; and solving the rigidity optimization comprehensive model to acquire an optimal solution to a rigidity design variable. Through the robot rigidity design method, an optimal distribution scheme of robot rigidity parameters can be found under the robot tail-end rigidity constraint; the robot rigidity and thestructure design can be more reasonable; and the robot rigidity design method has important significance in reducing the robot manufacturing cost.

The invention provides a flight scheduling method and device and equipment thereof. The method comprises the steps of obtaining an object loss value according to feature information of flight scheduling; determining a flight adjustment target value according to the object loss value; determining flight adjustment data according to the flight adjustment target value; and generating a flight scheduling plan according to the flight adjustment data. According to the technical scheme, the flight scheduling plan can be automatically generated, business personnel do not need to manually generate theflight scheduling plan, and the flight scheduling plan is quickly generated.

The invention provides a method for predicating voltage of a solar unmanned aerial vehicle cell panel. The method for predicating the voltage of the solar unmanned aerial vehicle cell panel comprisesthe following steps of 1, acquiring an operation parameter of the solar cell panel, obtaining a training set and a prediction set according to the operation parameter, and performing normalization processing on operation parameter data of the solar cell panel in a sunny day; 2, building a BP neural network according to the normalized training set; 3, obtaining the initial weight and initial threshold value of the BP neural network through optimization according to an adaptive differential algorithm; 4, using the optimized weight and threshold value as the initial weight and initial threshold value of the BP neural network, and training the BP neural network by means of the training set; and 5, inputting the to-be-predicated data in the predication set into the BP neural network, comparingthe predication value which is output from the BP neural network with an actual value, and evaluating accuracy of a predication result.

An embodiment of the invention relates to the technical field of information processing, and discloses an order allocation method, an order allocation device, a server and a nonvolatile storage medium. The order allocation method comprises the steps of: acquiring combinations of at least two to-be-allocated orders and candidate delivery resources according to the to-be-allocated orders and the candidate delivery resources, wherein each combination comprises one to-be-distributed order and one candidate distribution resource; calculating the distribution cost corresponding to each combination;acquiring an allocation result of each to-be-allocated order according to a preset constraint condition and the distribution cost corresponding to each combination, wherein the combinations correspondto at least one first target feature, and the constraint conditions include that for each first target feature, the sum of eigenvalues of the first target features of each combination meets a presetcondition. According to the order allocation method and the order allocation device provided by the embodiment of the invention, a more reasonable order allocation scheme can be provided, so that moreactual business scenes are covered.

The embodiment of the invention provides a resource allocation optimization method and device, and the method comprises the steps: obtaining an unscheduled process corresponding to the earliest starting moment according to the pre-obtained starting moments of all unscheduled processes in an unscheduled process set of a target project, and taking the unscheduled process corresponding to the earliest starting moment as a target process; According to various resources required for completing each target procedure and the required resources available at the earliest starting moment, allocating theavailable required resources to each target procedure based on a genetic algorithm, so as to minimize the complexity of a resource flow network generated by current allocation; And removing the target procedure from the unscheduled procedure set, and iteratively executing the steps of obtaining the target procedure and resource allocation according to the unscheduled procedure set from which thetarget procedure is removed until the unscheduled procedure set is empty. According to the embodiment of the invention, global optimization of resource configuration is realized, and the resource flownetwork structure is simplified.