263 results about "Mutation (genetic algorithm)" patented technology

The invention discloses a UUV cluster dynamic task planning method based on a multi-target genetic algorithm, a program, equipment and a storage medium, and belongs to the field of underwater multi-UUV cooperative detection of multiple targets. According to the method, firstly, for path planning of regional task points, chromosome representation is completed through sequential coding, chromosomes are selected and ranked randomly, and the chromosome with the highest fitness value is selected from each group; then, in each group of the current population, selecting an optimal parent individual to carry out crossover and mutation operations so as to improve the quality of offspring individuals, and carrying out updating to obtain a next-generation population; and finally, obtaining a plurality of optimal individuals, and outputting a task planning path result. Dividing the search area into a plurality of task sub-areas, and completing the dynamic task planning of the UUV cluster according to the target distribution condition of the sub-areas and the condition of each UUV. According to the method, the search strategy can be adjusted in real time according to the real-time detection result and the environment change, so that the target distribution non-uniformity and the cluster efficiency difference are effectively relieved.

The invention provides a workshop scheduling method and system fusing a decision tree and a genetic algorithm. The method comprises the following steps: firstly, collecting operation process and machine information through an ERP system and extracting related features; generating a scheduling scheme by using historical orders of an ERP system or manually added orders, and constructing a training set training decision tree to accurately judge machine allocation conflicts; constructing a multi-target flexible job shop scheduling model, and constructing a target function based on a hierarchical Pareto dominance relationship; and secondly, realizing job-shop scheduling scheme coding by adopting double-layer chromosome coding, carrying out selection, intersection and mutation operations in combination with a genetic algorithm, carrying out conflict detection and repair by utilizing a decision tree, solving an objective function, continuously iterating until a convergence condition is met, and outputting a Pareto optimal solution. According to the workshop scheduling method, the scheduling efficiency and feasibility are improved by establishing a data model of a multi-target flexible job workshop scheduling problem and through dynamic conflict detection, hierarchical multi-target optimization and a closed-loop feedback mechanism.

The invention relates to the technical field of water conservancy projects, in particular to a construction scheduling method for collaborative linkage of cross-basin water conservancy projects, which comprises the following steps of: constructing a space-time water conservancy topological network to realize multi-source data fusion and spatial association and provide a basis for accurate scheduling; potential contradictions between construction and hydrological regulation and control are found in advance by recognizing conflict areas and performing quantitative analysis; a space-time safety window strategy set is formulated, the construction period is optimized in combination with risk prediction, and operation safety and efficiency are guaranteed; through task decomposition and collaborative factor allocation, an atomic operation unit and a dependency relationship are defined, and the cross-project linkage capability is improved; by generating a preliminary scheduling scheme, and through a resource-collaborative double-constraint graph and a genetic algorithm, efficient resource configuration and conflict minimization are realized; through dynamic risk monitoring and scheduling scheme reconstruction, hydrological abrupt change and construction deviation are quickly responded, cooperative factors are triggered for rebinding, and scheme adaptability and robustness are ensured.

The invention provides a battery pack thermal management system control method based on an improved adaptive genetic algorithm, and the method comprises the steps: building a battery pack thermal model considering the thermal coupling effect between battery modules, and constructing a dynamic change model of the temperature of the battery modules, obtaining temperature models of head and tail battery modules in the battery pack according to the dynamic change model of the temperature of the battery modules; constructing a multi-objective optimization function which comprehensively considers the energy consumption of the cooling system, the temperature control precision of the battery pack and the temperature uniformity of the battery module; an improved adaptive genetic algorithm is obtained by introducing an adaptive crossover variation mechanism, an elitist retention strategy and a dynamic penalty function; based on an improved adaptive genetic algorithm, solving the multi-objective optimization function to obtain an optimal coolant flow rate; the optimal cooling liquid flow speed is converted into a control instruction of the electronic water pump rotating speed through the water pump characteristic curve, and real-time adjustment of the electronic water pump rotating speed is achieved. According to the invention, accurate control of the battery pack thermal management system can be realized.

The invention relates to the technical field of electric power detection, in particular to an electric power inspection robot detection method and system based on machine vision. According to the method, the three-dimensional model of the target area is constructed by pre-collecting the basic data of the area, the historical defect data is obtained to mark the detection key area, the initial path and the key detection point are optimized by using the genetic algorithm, redundant collection points are removed, and the accuracy of the visual detection path is improved; meanwhile, visual data are analyzed in real time through a path acquisition collaborative analysis module, an optimization instruction is generated to dynamically adjust a path, and the problems of low efficiency and missing detection caused by independent operation of vision and navigation are avoided; the visual perception module collects detection dynamic data of definition, vibration acceleration, environment wind speed and environment light, the environment adaptability module analyzes and outputs a comprehensive fluctuation value, a corresponding fluctuation value interval is obtained, a corresponding environment adaptability regulation and control instruction is generated, and influences caused by motion blur and environment sudden change are avoided.

The invention relates to a visual-language model cue word evolution generation method based on a genetic algorithm, which comprises the following steps of: randomly generating different cue words under a target task to construct an initial cue word set, preprocessing the randomly generated cue words, and reserving N groups of cue words as an initial population; and designing a multi-dimensional fitness evaluation function to evaluate the performance quality of each group of cue words in a vision-language task, and selecting high-quality cue word individuals based on an elitism strategy. Performing crossover and mutation operation on the selected high-quality cue word individuals by referring to a genetic algorithm to generate a new-generation cue word population; and finally, carrying out iterative optimization for multiple times until a preset termination condition is met, and outputting an optimal cue word set. By the adoption of the visual-language model cue word evolution generation method based on the genetic algorithm, high-quality and diversified cue words can be automatically generated, the performance of a visual-language model is remarkably improved, and meanwhile the visual-language model cue word evolution generation method has good interpretability and adaptability.

The invention relates to a comprehensive energy system optimization scheduling method based on an improved particle swarm algorithm, and belongs to the technical field of comprehensive energy system optimization scheduling. The method comprises the steps of constructing an operation model of an integrated energy system with electrical cold and heat as main energy flows, constructing double-target integrated energy system optimization scheduling with the lowest operation cost and the optimal power supply reliability, and solving a multi-target problem through an improved particle swarm algorithm on the premise of meeting power balance constraints and equipment operation constraints. Self-adaptive variation and variable inertia factors are introduced on the basis of a traditional particle swarm algorithm, the self-adaptive variation refers to a variation thought in a genetic algorithm, and variation operation expands a population search space which is continuously reduced in iteration, so that particles can jump out of the position of an optimal value which is searched previously, search is carried out in a larger space, and the search efficiency is improved. The population diversity is maintained, and the possibility of searching the optimal value by the algorithm is improved. And the comprehensive energy system can give full play to economy and reliability under constraint conditions.

The invention belongs to the technical field of indoor positioning, and discloses a Bluetooth path loss model parameter adaptive calibration method and system. According to the method, the historical learning mechanism and the hybrid optimization strategy are fused, so that the global search capability and the local optimization precision are considered while the dynamic updating of the path loss model parameters is realized. Compared with an existing calibration method of a fixed parameter or a single optimization algorithm, the method can automatically correct the model parameters according to the environment change, and solves the problem that the precision of a traditional model is reduced under the conditions of multipath effect, shielding interference and environment sudden change. A hybrid optimization framework of a genetic algorithm and a particle swarm algorithm is introduced, so that a parameter optimization process obtains a high-quality initial value in a global search stage, rapid convergence is realized in a local fine adjustment stage, and the calibration efficiency and precision are remarkably improved. And meanwhile, a historical learning mechanism is added, so that the model has a time memory characteristic, smooth mutation of historical parameter information can be fused, and the stability and robustness of the method in a complex dynamic environment are improved.

The invention provides a redundant backup method and system for a communication bus in track type gravity energy storage, and relates to the technical field of intelligent control, and the method comprises the steps: carrying out the dynamic optimization of decision parameters in a fuzzy logic decision mechanism through employing a genetic algorithm based on an obtained original link state data set, and generating an optimized decision parameter set; according to the genetic algorithm, historical fault state data and a corresponding final switching decision are used as training samples, and decision parameters are iteratively optimized through selection, crossover and mutation operations; and based on the generated optimized decision parameter set, performing fuzzy logic decision processing on the currently collected state data of the main and standby links, and when the health state evaluation value of the main link is lower than a preset threshold value, triggering a millisecond switching instruction. According to the invention, fault processing of the communication link can be completed without manual intervention, and the intelligent level of the system is improved.

The invention discloses an electronic circuit energy efficiency optimization design method based on a genetic algorithm, and the method comprises the following steps: S1, carrying out the modeling of a to-be-optimized electronic circuit, forming a mixed type chromosome, and initializing a genetic algorithm population; s2, constructing a fitness function based on a genetic algorithm population; s3, executing a selection operation on the current population, and replacing a parent by a child with high adaptability; s4, performing crossover operation on the parent individuals to generate new offspring individuals; s5, performing mutation operation on the offspring individuals to form new individuals; s6, performing circuit simulation on the new individual after the genetic manipulation is executed, and inputting a simulation result into a fitness function for evaluation; s7, updating the population according to the fitness score, and entering next-generation evolution; and S8, configuring an individual parameter which meets a convergence condition and has the highest output fitness score. Based on a local competition algorithm, a double-layer block recombination strategy and a Pareto frontier algorithm are fused, and energy efficiency optimization of the electronic circuit is achieved.

The invention relates to the technical field of intelligent breeding and logistics optimization control, and solves the technical problems of high energy consumption, unstable efficiency, unbalanced distribution, lack of an intelligent scheduling mechanism and the like in feed conveying of a multi-storey pig farm. The method comprises the following steps: acquiring static parameters (physical characteristics of feed, physical attributes of a conveying system and a pig house structure) and dynamic parameters (real-time feeding requirements, equipment and material states and external environment factors); establishing a multi-objective optimization model of a collaborative optimization energy consumption model E (x), a time model T (x) and a conveying balance degree model U (x); solving by adopting a genetic algorithm with a special design crossover and mutation operator to obtain an optimal scheduling scheme; in the execution process, a closed-loop self-learning calibration mechanism is started, actual power is measured through a current sensor, and when the deviation between predicted energy consumption and actual energy consumption exceeds a preset threshold value, efficiency parameters in the energy consumption model are reversely corrected through a gradient descent method; the weight coefficients of the three models are dynamically adjusted according to the real-time electricity price and the inventory state. The system adopts a three-layer architecture of a perception and data acquisition layer, a decision and control core layer and an execution layer. According to the method, multi-target collaborative optimization and intelligent adaptive scheduling are realized, the total energy consumption is effectively reduced, the transmission time is shortened, the distribution balance degree is improved, and the energy consumption prediction accuracy is remarkably improved.

The method comprises the following steps: obtaining a power quality signal containing noise; selecting permutation entropy as an adaptive function of genetic algorithm, calling variational mode decomposition through genetic algorithm, and iteratively optimizing a penalty factor α and a decomposition mode number k of the variational mode decomposition to determine optimal parameters; decomposing signal data into k mode components through the variational mode decomposition, and determining effective mode components and noise mode components through a correlation coefficient; for improved wavelet threshold, a parameter-adjustable threshold function is proposed, and the concept of wavelet energy entropy is introduced into the threshold function; the noise mode components are denoised through the improved wavelet threshold, and the effective mode components and the denoised noise mode components are reconstructed to obtain a denoised power quality disturbance signal. The method can effectively remove noise interference while retaining singular information of mutation points of the collected signal, and provides help for subsequent analysis and treatment of the power quality disturbance signal.

The invention discloses an underwater multi-target routing method based on an improved non-dominated sorting genetic algorithm. The method comprises the following steps: firstly, arranging sensor nodes in a target water area, networking, and encoding a feasible path from a source node to a target node by adopting a segmented structure; secondly, constructing a routing multi-target fitness function, and introducing a dynamic weighting function to form a comprehensive fitness index; and finally, grading the population by using an improved non-dominated sorting genetic algorithm, and exploring a better path through crossover and mutation operations. And designing a reference point elite selection strategy, improving the coverage of the solution set in the target space, and outputting an optimal routing path until the algorithm converges. According to the method, the dynamic balance among different optimization targets can be realized, and the link quality is considered while the energy consumption and the time delay are reduced, so that the network routing performance is improved.

The invention relates to a low-temperature high-strength steel lining polytetrafluoroethylene material based on a novel plasticizer and a preparation method of the low-temperature high-strength steel lining polytetrafluoroethylene material, and belongs to the field of anticorrosive materials. The method comprises the following steps: constructing a multi-dimensional associated training set by taking a historical plasticizer component ratio, tensile strength, elongation at break and temperature resistance indexes of a material and sizes and functional parameters of different products as input; by defining a multi-objective fitness function, synthesizing the material strength, low-temperature toughness and cost economy, utilizing the global search capability of genetic algorithm selection, crossover and mutation operation, and combining with a simulated annealing algorithm, the local optimization characteristic of collaborative iterative optimization is realized by dynamically adjusting the inferior solution accepting probability, and the optimal proportioning scheme of the plasticizer is gradually approached. According to the SA algorithm, global exploration and local development are balanced through a temperature attenuation mechanism, the GA premature convergence problem is effectively avoided, the robustness of a matching scheme is improved, and the efficiency and precision of material performance optimization are remarkably improved.

The invention provides a discrete mixed job production line scheduling method based on an improved genetic algorithm, and relates to the technical field of industrial automation and production scheduling, and the method comprises the steps: S1, carrying out the modeling and data input of a production line scheduling problem, constructing a flexible job shop scheduling model, defining a decision variable, a constraint condition and a target function, and inputting basic data; s2, generating an initial population by adopting a hybrid initialization strategy, wherein the initial population comprises randomly generated individuals and individuals generated based on a heuristic rule; s3, chromosome coding is carried out on the scheduling scheme in a two-segment coding mode, wherein a process sorting segment and a machine distribution segment are included; s4, the fitness is calculated, and individual selection is carried out by adopting a tournament selection method; s5, executing improved genetic operations including adaptive crossover and mutation operations; s6, carrying out local search on the elite individuals, wherein the local search comprises key path identification and neighborhood disturbance; and S7, judging a termination condition, if the termination condition is met, outputting an optimal scheduling scheme, otherwise, returning to the step S4.

The invention discloses a large model output data security detection method and system based on adversarial attacks. The method comprises the following steps: constructing and optimizing a strategy space containing a plurality of attack strategies, and grading and sorting the strategies to improve the attack efficiency; generating a single-strategy antagonism prompt by the attack model according to the optimized strategy space, and performing effectiveness evaluation and feedback correction on the prompt by the judgment model; inputting a prompt passing the evaluation into the target large model to obtain a response, and performing malicious degree scoring on the response by the judgment model; and if the single-strategy attack is not successful, introducing an optimization mechanism based on a genetic algorithm, generating a more complex multi-strategy antagonism prompt through strategy variation and crossover, and carrying out iterative attack until the target large model is successfully broken into the prison. According to the method, the security defects of the large model can be efficiently and comprehensively detected in a self-adaptive and multi-strategy attack mode.

The invention provides an unmanned aerial vehicle cluster weapon target collaborative allocation method and system under space-time constraint, and the method comprises the steps: defining a collaborative allocation object, constructing a collaborative combat scene target function, defining a collaborative combat scene constraint condition, and finally solving the target function through employing a multi-gene population parallel ant colony algorithm, encoding the weapon set, the target set and the unmanned aerial vehicle set into a weapon gene sequence, a target gene sequence and an unmanned aerial vehicle gene sequence respectively to generate an initial population, and performing staged joint optimization by fusing a pheromone guiding mechanism of an ant colony algorithm and crossover mutation operation of a genetic algorithm, the optimal cooperative allocation scheme is searched when the constraint condition is satisfied, the unmanned aerial vehicle task allocation result can be better obtained and optimized through the cooperative allocation mode, and the unmanned aerial vehicle cooperative allocation efficiency is improved.

The invention discloses a game optimization solving method for a multi-user power control problem in a wireless network, and the method comprises the following steps: building a multi-user power control game model, and arranging a joint strategy vector x according to the transmission power decision of a user; designing a fitness function F (x); selecting an elite individual with F (x) closest to 0 in the genetic algorithm population; and updating a local search radius by adopting a cosine annealing algorithm, performing optimization comparison on elite individuals in combination with a particle swarm optimization algorithm, forming a new generation of population in combination with new individuals obtained by selection, crossover and variation, and outputting an individual with F (x) closest to 0 after iteration, namely a Nash equilibrium solution as a final power control decision. According to the method, a game problem is converted into an optimization problem by designing a specific fitness function, so that the calculation complexity and the implementation threshold are remarkably reduced; and the method has low requirements on the property of a cost function, does not need to be continuous or differentiable, has relatively high robustness and universality, and can efficiently solve a continuous game problem.

The invention discloses an internal combustion engine performance multi-objective optimization method based on data driving and a genetic algorithm, and belongs to the technical field of internal combustion engines. The method comprises the following steps that performance data of power, combustion, emission and the like of the internal combustion engine under different working conditions are collected, and an internal combustion engine performance data set is constructed; training a machine learning model based on the data set, and establishing an internal combustion engine performance prediction model; a genetic algorithm is adopted, an initial population is generated within a preset internal combustion engine working condition range, and fitness values of population individuals are rapidly calculated with the help of the internal combustion engine performance prediction model; and continuously iteratively optimizing the population through the steps of intra-population selection, crossover, variation, fitness calculation, evaluation and the like until the population meets an expected convergence condition or reaches the maximum number of iterations, thereby obtaining a Pareto optimal solution set for multi-objective optimization of the performance of the internal combustion engine. According to the method, rapid evaluation and multi-target performance optimization of the engine performance can be realized, and the number of experiments and the development period are greatly reduced.

The invention relates to an islet repair ICK mode oral polypeptide optimization system taking VDAC1 as a target based on ESM2 and a genetic algorithm, and the system comprises a mutation module which is used for carrying out mutation and recombination by adopting the genetic algorithm according to original protein sequence data to generate an initial population; the prediction module is used for respectively inputting the protein sequence data of the initial population into a protein dimer dissociation energy prediction model, a protein toxicity prediction model and a protein cell penetrability prediction model to obtain a dissociation energy prediction value delta G, a toxicity probability value T and a cell membrane passing transmission probability Ed; and the optimization module is used for acquiring the optimal individual performance of the highest fitness score to output the Pareto optimal sequence index, and recording the hotspot residue distribution and the site mutation frequency at the same time. On the basis of keeping the biological activity of the original sequence, the key drug properties are improved in a breakthrough manner.

The invention provides a cooperative path planning method for searching and tracking an underwater target by a multi-USV unmanned cluster, and the method comprises the steps: introducing a pose-control quantity-time to carry out the three-dimensional double-chain coding of a population in an improved genetic algorithm, enabling a control instruction and an accurate execution timestamp to serve as a genetic gene, and enabling the control instruction and the precise execution timestamp to serve as a genetic gene in the crossover and mutation operation, timestamps, control quantities and poses are transmitted synchronously, and it is ensured that offspring individuals can inherit an excellent cooperation mode of a parent; during population initialization, USV is guided to preferentially cover a high-value region through region division and weighted Gaussian distribution, so that blindness caused by random initialization of a traditional genetic algorithm is overcome, and invalid search is avoided; a target for avoiding secondary search is designed in a fitness function, and repeated access to recently searched areas is quantified by introducing a concept of confidence time intervals, so that an algorithm actively explores unexplored water areas, waste of search resources is avoided, and accuracy and planning efficiency of multi-USV collaborative path planning are improved.

The invention relates to the technical field of data processing, in particular to a polycaprolactone polyol synthesis path optimization method based on a genetic algorithm, which comprises the following steps: constructing an antagonism framework taking the genetic algorithm as a generator and a process stability discriminator; directionally generating virtual failure data by using a failure path deduction model so as to solve the problem of data sparsity and train a discriminator; quantifying the robustness risk of the candidate process parameters through a virtual disturbance unit, and taking the risk as a key part of a fitness function of the genetic algorithm; meanwhile, the interpretability analysis result of the discriminator is used for guiding the variation direction of the genetic algorithm. According to the method, the process path with high performance and high robustness can be found, and the industrial practicability and decision-making efficiency of an optimization result are remarkably improved.

This invention proposes a GQA-LUT method, utilizing a genetic algorithm and LUT-based circuit to efficiently approximate non-linear operators in Transformers. It adaptively finds optimal solutions for various non-linear functions, outperforming conventional neural network methods. A novel rounding mutation (RM) algorithm enhances approximation accuracy during quantization, improving low-bit integer precision. The invention also introduces a LayerNorm folding strategy as a near-memory computing principle, reducing IO and energy overheads with a two-stage memory hierarchy. Additionally, an additive partial sum quantization method is proposed to reduce energy consumption by quantizing accumulated PSUMs in matrix multiplication, alongside a PSQ-APSQ grouping strategy and floating-point regularization.

The present application relates to a kind of neutron energy spectrum optimization method and system for improving reactor radioisotope production efficiency, method includes the following steps: initial population of genetic algorithm is randomly generated, these population individuals are all neutron energy spectrum divided into n energy group, the body average neutron energy spectrum in target piece when the neutron energy spectrum of each individual is calculated as irradiation hole incident spectrum, the radioisotope production under the body average neutron energy spectrum in target piece corresponding to each individual is calculated, the evolution of population is realized by the mutation and crossing operation of genetic algorithm, the above-mentioned process is repeatedly executed until iteration converges, obtains the neutron energy spectrum regulation scheme for realizing the maximization of radioisotope production.Compared with prior art, the present application efficiently realizes the accurate regulation of neutron energy spectrum in reactor, can provide technical support for the energy spectrum regulation process of radioisotope irradiation production in reactor.

The invention belongs to the technical field of production scheduling, and provides a genetic algorithm self-learning variation method and system for solving a scheduling problem, and the method comprises the steps: a chromosome initialization step: setting a workshop variable and a dynamic scheduling rule, and generating an initial solution through a greedy method heuristic strategy; and a chromosome iterative optimization step: iteratively optimizing the initial solution according to a genetic algorithm to obtain a variation result. According to the method, the GA is combined with the self-learning mutation method, and a single-task and batch-task moving mechanism of the same unit and cross units is adopted, so that self-adaptive adjustment of a chromosome structure is realized, the global optimization capability and scheduling efficiency of the algorithm are improved, the global search capability of the algorithm is enhanced, the convergence speed is increased, and the optimization result is improved; by replacing heuristic variation driven by traditional rules with self-learning variation, the complexity of customized development is reduced, the scheduling efficiency is improved, and meanwhile, the code quantity and the implementation cost are reduced.

The invention discloses an official vehicle scheduling and optimizing method and system for mixed type tasks, and belongs to the field of official vehicle intelligent scheduling. The method comprises the following steps: acquiring basic information of official business tasks, drivers and vehicles, constructing a task distance matrix, then establishing a scheduling model which takes the minimum total fuel consumption as a target and contains constraints such as grade matching and passenger-carrying upper limit, and setting genetic and simulated annealing algorithm parameters; constructing an initial scheme population through chromosome codes matched with task levels, and calculating the fitness value of each scheme; and taking the chromosome as a simulated annealing initial solution to complete local optimization, transmitting the solution into a genetic algorithm to carry out crossover and mutation iteration, and finally outputting an optimal scheduling scheme. According to the method, a hybrid algorithm framework is adopted, global search and local optimization capabilities are considered, the compliance, low oil consumption and high resource utilization rate of a scheduling scheme are realized, the problems of level mismatching, resource waste and algorithm limitation of traditional scheduling are solved, and the method is suitable for official car scheduling scenes of various organizations, enterprises and public institutions.

This invention provides a distributed generation (DG) optimization configuration method based on a quantum genetic algorithm, aiming to minimize power loss and improve voltage stability in power distribution systems. The method first collects basic data of the power distribution network, including bus numbers, voltage levels, line impedances, bus load demands, and capacity limitations of distributed generation sources. Next, a population containing different DG configuration schemes is initialized using a random number generation algorithm, and the population is iteratively optimized using a quantum genetic algorithm. During the optimization process, quantum selection, crossover, and mutation operations are used to generate a new generation of configuration schemes, and the power loss and voltage level of each scheme are evaluated using a fitness function to gradually approach the optimal solution. After obtaining the optimal configuration scheme, different load conditions (such as load increases or decreases) are simulated to verify and optimize the adaptability of the configuration, ensuring that the system can maintain low power loss and voltage stability under load fluctuations. The final output optimal DG configuration scheme includes bus location, capacity, and optimized power loss and voltage curves, capable of adapting to dynamic load changes. This method can effectively reduce power loss and improve voltage stability in power distribution systems, providing technical support for building efficient and reliable power distribution systems.

The application provides a mixed-band controlled SDN controller deployment method and device, wherein the method comprises the following steps: taking network delay, load difference and bootstrap time of an SDN system as optimization targets, taking cost as a constraint condition, and determining a target function; based on a third generation non-dominated sorting genetic algorithm, a feasible solution construction algorithm, a crossover mutation algorithm, a removal of isolated node algorithm, and a random intervention and update reference point algorithm, the target function is solved to determine an optimal solution of a Pareto frontier; and the optimal solution of the Pareto frontier is taken as a controller deployment scheme of the SDN system, and the controller of the SDN system is deployed. Thus, the SDN controller deployment problem under the condition of multi-objective optimization can be effectively optimized, and the requirements of low delay, load balancing and the like can be met.

The invention relates to a reinforcement learning-based river-lake-underground water combined dispatching optimization method, which specifically comprises the following steps of: performing space-time normalization on river-lake-underground water combined dispatching data, and defining a space-time feature mapping function to complete mapping of the data to a new feature space; constructing a joint scheduling model based on the mapping output, and defining a state space, an action space and a reward function to obtain a state vector, an action vector and a reward value; optimizing a genetic algorithm, constructing a reinforcement learning agent in combination with an actor-commentator architecture, initializing a network weight and a population by using state space clustering information, screening elite individuals by calculating individual fitness, fusing multiple features to realize adaptive crossover variation to update the population, synchronously updating an actor network and an action vector, and setting a re-learning mechanism; and after strategy fusion update and action execution are completed, final optimization of joint scheduling is realized based on the optimal strategy parameter and the network update weight. According to the invention, the optimization performance and adaptability of the scheduling strategy are improved.

This invention discloses a method, apparatus, medium, and device for parameter optimization of analog circuits, relating to the field of analog circuit technology. The invention automatically identifies current paths through a search algorithm to deeply analyze the circuit topology, determine the symmetry structure in the analog circuit, and constructs symmetry constraints based on the identical optimization parameters of electrical components within the symmetry structure. Simultaneously, it transforms the product of the transistor's gate width and the number of parallel connections into the total width and sets the optimization boundary. This combination reduces the dimensionality of the optimization variables while ensuring that the parameter optimization process conforms to the layout design of the analog circuit. Furthermore, it combines a population mutation algorithm and a non-dominated sorting genetic algorithm with an elitist strategy to perform iterative optimization within the constructed efficient search space, significantly improving the convergence speed of analog circuit parameter optimization, reducing computational costs, and enhancing the practicality of the analog circuit parameter optimization method.