25 results about "Gaussian mutation" patented technology

The invention discloses a tractor transportation operation working condition construction method based on improved particle swarm optimization and KMeans fusion, and relates to the technical field of agricultural machinery working condition analysis. The method comprises the following steps: acquiring original data of tractor transportation operation through a plurality of data acquisition modes, and carrying out preprocessing and three-stage screening to obtain an effective kinematics fragment; selecting multi-dimensional characteristic parameters to construct a characteristic matrix, and performing data dimension reduction by adopting principal component analysis; optimizing a KMeans clustering initial center by using an improved particle swarm optimization (IPSO) algorithm which introduces a dynamic inertia weight and a Gaussian mutation strategy, and performing clustering analysis on the feature space after dimension reduction; and selecting representative fragments based on feature similarity, and synthesizing a standardized working condition curve by taking the sum of average relative errors of all feature dimensions as a target function. According to the method, the problems that a traditional clustering algorithm is prone to falling into local optimum and the working condition construction accuracy is insufficient are solved, the constructed working condition can truly and comprehensively reflect the actual transportation operation characteristics of the tractor, and a reliable basis is provided for tractor power system optimization, operation efficiency improvement and energy consumption reduction.

The invention discloses a permanent magnet synchronous motor parameter optimization method based on an improved moth flame particle swarm algorithm. In order to solve the problems that a traditional PID controller is difficult in parameter setting, poor in adaptability, insufficient in control precision and the like under complex working conditions, in the strategy, an MFPSO algorithm fuses a particle swarm optimization (PSO) algorithm and a moth flame optimization (MFO) algorithm. An adaptive inertia weight and Gaussian mutation strategy is introduced, so that particles can focus on local fine search, and the speed of convergence to an optimal solution is accelerated. The optimization capacity of the algorithm in different stages is enhanced by adopting a nonlinear acceleration coefficient, and local optimum is effectively avoided. A grouping evolution mechanism is designed to promote intra-group information sharing and inter-group competition, the population diversity is enriched, and the ability of the algorithm to jump out of a local optimal solution is further improved. The fuzzy PID controller based on the MFPSO algorithm optimization parameters can better meet the requirements of high-performance control of permanent magnet synchronous motors in different fields, and has application prospects and economic values.

The invention provides a curved surface circuit jet printing droplet spreading prediction method for optimizing a BP neural network (GA-BP) based on a genetic algorithm, and belongs to the technical field of inkjet printing process parameter optimization and curved surface electronic manufacturing. A high-precision prediction model is constructed for solving the problem that spreading behaviors of micro-droplets on a cylindrical surface substrate are complex due to coupling influences of material physical properties, substrate characteristics and process parameters. The dynamic viscosity, the surface tension, the density, the surface wettability, the curvature radius, the impact speed and the temperature serve as input, the circumferential and axial maximum spreading diameters serve as output, and training data are obtained through high-speed camera shooting or CFD simulation; globally optimizing the initial weight and threshold of the BP network by adopting a genetic algorithm, and simulating binary crossover and Gaussian mutation strategies by taking the reciprocal of a verification set MSE as a fitness function and combining roulette selection; finally, the GA-BP model obtained through training meets the conditions that R2 is larger than or equal to 0.92 and RMSE is smaller than or equal to 0.05, and real-time prediction of the spreading behavior under the new process condition can be achieved. The method effectively overcomes the problems that an empirical formula is poor in applicability and low in numerical simulation efficiency, and traditional BP is prone to sinking and local minimum and the like, and has high precision, strong generalization and engineering practicability.

This invention discloses a multi-threshold image segmentation system, method, and medium for supply chain product inspection. The multi-threshold image segmentation method for supply chain product inspection includes: acquiring supply chain product images; preprocessing the supply chain product images to obtain preprocessed images; generating explosion sparks based on the preprocessed images using an explosion operator; randomly selecting several fireworks and performing Gaussian mutation operations to generate mutated sparks; mapping explosion sparks and mutated sparks exceeding the boundary range back to the feasible range according to an exponential perturbation mapping rule, and generating a new fireworks population according to a random selection strategy; determining whether the new fireworks population meets the termination condition; if so, segmenting the preprocessed images using the optimal threshold to obtain segmented images; and detecting supply chain products based on the segmented images. This multi-threshold image segmentation method for supply chain product inspection solves the problem in existing technologies where the human eye cannot easily identify slightly damaged parts of products in the equipment supply chain.

This invention proposes a method for computing offloading and resource allocation in the Internet of Vehicles (IoV) based on an improved Black Kite optimization algorithm. The method comprises the following steps: obtaining the computational latency and computational energy consumption of tasks in various computing scenarios based on multiple computing scenarios; calculating the total computational latency and total computational energy consumption of the tasks based on the computational latency and computational energy consumption of the tasks in various computing scenarios; defining a system utility function based on the total computational latency and total computational energy consumption of the tasks; constructing a computational latency and energy consumption model for mobile edge computing scenarios with the goal of minimizing the system utility function; designing a Black Kite optimization algorithm based on an elite reverse learning strategy, a Gompertz model, and a Gaussian mutation strategy; and using the designed Black Kite optimization algorithm to solve the computational latency and energy consumption model for mobile edge computing scenarios to obtain a final computation offloading and resource optimization allocation solution. This invention can accelerate the convergence of the algorithm and prevent the algorithm from prematurely falling into a local optimal solution.

The application relates to a WSN cluster head election method and system based on an elite multi-objective gold panning algorithm, and the method comprises the following steps: calculating an optimal cluster head number based on a wireless sensor network; generating an initial population by using a Sobol sequence and performing population iteration; updating individuals of the population by using an optimized gold panning optimization algorithm to obtain a first population; performing Gaussian mutation on the first population to obtain a second population; combining the three populations to form an enhanced elite candidate population; performing target function calculation, quickly non-dominant sorting the individuals by using an NSGA-III algorithm to obtain the level to which the individuals belong; judging whether the current iteration number reaches a target iteration number; if not, selecting the next generation population based on the enhanced elite candidate population and by using the NSGA-III algorithm to perform population iteration; and if yes, outputting a Pareto front to determine a cluster head node in a wireless sensor network (WSN). The application introduces the enhanced elite candidate population to expand population diversity and avoid loss of historical elite solutions.

This invention provides a collaborative path planning method for underwater target search and tracking based on AUVs and multiple USVs. It employs a distributed, cross-domain, heterogeneous collaborative path planning architecture. By setting different fitness functions in the search and tracking phases, the USV cluster and AUVs can dynamically adjust their path planning strategies according to mission requirements, achieving close collaboration between cross-domain platforms within the same time period and overcoming the information transmission lag problem in traditional phased collaborative modes. In the tracking phase, this invention introduces a dynamic periodic underwater target position prediction method to predict the target's future trajectory, enabling the AUV to switch from tracking mode to interception mode, improving the target acquisition success rate. The algorithm effectively balances global exploration capability and local development accuracy by introducing adaptive explosion radius, Levy flight strategy, and Gaussian mutation operation, resulting in stronger global optimization capability and faster convergence speed in the optimization of multi-USV and AUV collaborative path planning.

The invention discloses an intracranial hemorrhage CT image segmentation method and device, a medium and equipment, and relates to the technical field of image segmentation. In a segmentation threshold optimization process, a Bareones mechanism and an orthogonal learning mechanism are integrated into an African vulture optimization algorithm, a search space is expanded in a search population corresponding to a segmentation threshold through Gaussian mutation, and the ability of the threshold optimization algorithm to get rid of local optimum is enhanced; and an orthogonal learning mechanism promotes resource sharing among search population individuals corresponding to the segmentation threshold, and improves the search diversity of the segmentation threshold. And obtaining an optimal individual with the maximum Renyi entropy after the two-dimensional histogram is divided based on the threshold value through multi-round iterative optimization, and taking the optimal individual as an optimal segmentation threshold value of the intracranial hemorrhage CT image. According to the method, the population diversity and convergence speed in segmentation threshold optimization of the intracranial hemorrhage CT image are improved, and the image segmentation effect of the intracranial hemorrhage CT image is improved.

The invention provides an improved competitive particle swarm optimization (CSO) algorithm in order to solve the problems of load unevenness and task completion time optimization in cloud computing task scheduling. Aiming at the problems of slow convergence, easy falling into local optimum, neglect of load balance and the like when the traditional CSO is used for processing the problems, three innovations are carried out: 1, the global exploration capability of the algorithm is effectively enhanced by introducing Gaussian mutation operation to implement position disturbance on winner particles; 2, designing a dual-mode updating strategy, and guiding the population to escape from a local optimal region in an evolution stagnation stage through a differentiated particle updating mechanism; and 3, constructing a multi-objective fitness function, and integrating task completion time (Makespan) and a load balancing index (LB) to realize collaborative optimization of scheduling efficiency and resource utilization. By reconstructing an updating mechanism of the competitive particle swarm, the comprehensive performance of the algorithm in the aspects of scheduling performance and resource utilization rate is remarkably improved, and a more competitive solution thought is provided for the task scheduling problem in the cloud computing environment.

The invention discloses a motor bearing fault diagnosis method and system based on FMD and Transform, belongs to the technical field of motor bearing fault diagnosis, solves the problem of how to improve the fault diagnosis precision of a motor bearing, and obtains MBKA by optimizing BKA through Funch chaotic mapping, a golden sine optimization strategy and a Gaussian mutation strategy, and the algorithm has higher optimization capacity. Parameters of the FMD and the Transform are optimized through the MBKA, so that the uncertainty of manual parameter setting is avoided, and the preprocessing effect of the signal and the fault diagnosis performance of the model are improved. In combination with a motor bearing fault mechanism, the influence of a strong noise environment on a fault diagnosis result is reduced, and feature information in vibration data of different fault types of the motor bearing is fully extracted, so that fault diagnosis is accurately carried out on the motor bearing, and personnel and equipment damage is avoided.

The application discloses a deep optical neural network training method and system based on a hybrid mutation strategy genetic algorithm, and the method comprises the following steps: S1, sequentially stacking a linear operation layer based on MZIs, a nonlinear activation layer based on EOA and a Dropmask based on a mask to build an N-layer deep DONN; S2, preprocessing a data set with different characteristic categories to conform to the data input size of the DONN; S3, uniformly initializing the DONN population, combining the MSE and the Accuracy between the real value and the predicted value as the fitness evaluation function of the individual; S4, taking the exponential ranking selection ERS and the uniform crossover UC as the selection operator and the crossover operator in the training process, adopting a hybrid mutation strategy, and distributing three operators, namely, the single-point mutation SM, the uniform mutation UM and the Gaussian mutation GM, to different individuals for mutation according to a dynamic game probability; and S5, adopting a double-elite reservation strategy, reserving two individuals with the optimal MSE and Accuracy performance to the next generation, and through iterative evolution, until a termination condition is met, and a DONN individual with the globally optimal network parameter is obtained.

The invention relates to the technical field of optimal solution search, in particular to a genetic model-based artificial intelligence optimal solution search method, which comprises the following steps of: generating an initial population by utilizing a Latin hypercube sampling method according to a constraint condition, uniformly distributing individuals in a solution space, calculating individual fitness values, constructing a fitness distribution histogram, and performing optimal solution search. Genotype and phenotype diversity indexes are monitored in real time, when the diversity indexes reach a specific threshold value or meet early warning conditions, gradient descent local search is performed on the optimal individual, a dynamic step length adjustment strategy is adopted to optimize the search process, validity verification is performed on an optimization vector, and when local search is not triggered, the optimal individual is subjected to gradient descent local search. According to the method, the hierarchical selection, single-point crossover and Gaussian mutation operations are performed on the population, and the boundary of a variation individual is repaired, so that the problems of insufficient population diversity control and easy local optimum in a complex solution space in a traditional method are solved, and the search efficiency and the result quality are improved.

The application discloses an intracranial hemorrhage CT image segmentation method and device, medium and equipment, and relates to the technical field of image segmentation. In the threshold segmentation optimization process, the Barebones mechanism and the orthogonal learning mechanism are integrated into the African vulture optimization algorithm. The search space is expanded in the search population corresponding to the segmentation threshold through Gaussian mutation, and the ability of the threshold optimization algorithm to get rid of local optimization is enhanced. The orthogonal learning mechanism promotes the resource sharing between the search population individuals corresponding to the segmentation threshold, and improves the search diversity of the segmentation threshold. Through multiple rounds of iteration optimization, the optimal individual that makes the Rényi entropy of the two-dimensional histogram based on the threshold segmentation maximum is obtained as the optimal segmentation threshold of the intracranial hemorrhage CT image. The application improves the population diversity and convergence speed in the threshold segmentation optimization of the intracranial hemorrhage CT image, and improves the image segmentation effect of the intracranial hemorrhage CT image.

The invention relates to a method and a system for realizing photovoltaic array fault diagnosis based on an integrated learning model CatBoost, which can solve the problem of low fault diagnosis accuracy of a photovoltaic array in a complex terrain and a changeable environment. An existing diagnosis method is limited by environmental factors and insufficient data, key information in a current-voltage characteristic curve (I-V curve) cannot be fully mined, and therefore fault type recognition is limited. In order to solve the problems, an I-V curve correction algorithm is firstly provided and used for correcting the influence of environment variables (such as temperature and irradiance) on fault feature representation, and feature information with higher identification capacity is extracted. Then, a CatBoost model is adopted to realize high-precision real-time fault diagnosis under a photovoltaic array small sample condition, and a sparrow search algorithm (SSA) is utilized to optimize key hyper-parameters of the model so as to improve the diagnosis performance and generalization ability of the model. Furthermore, in order to enhance the optimization capacity of the sparrow search algorithm, an improved sparrow search algorithm (EKSSA) fusing an elite reverse learning strategy and a Cauchy Gaussian mutation strategy is introduced, and the performance of the CatBoost model is optimized, so that the CatBoost model is more excellent in fault detection and classification. Through the method, real-time monitoring, intelligent fault diagnosis and predictive maintenance of the photovoltaic array under complex working conditions can be realized, the reliability and the operation efficiency of the system are remarkably improved, and the method is suitable for practical application scenes of distributed energy systems such as photovoltaic power stations and the like.

The application discloses an electric energy substitution potential prediction method and system based on ISSA-ELM, relates to the technical field of electric energy substitution potential prediction, and comprises the following steps: analyzing influence factors of electric energy substitution potential and quantifying, inputting quantized values of the influence factors of electric energy substitution potential; configuring parameters of a sparrow search algorithm, performing global and local search optimization based on a chicken swarm optimization algorithm and a Cauchy-Gaussian mutation strategy; evaluating a fitness function to obtain an optimal adaptive individual, calculating optimal weights and thresholds of an extreme learning machine, establishing an optimal extreme learning machine prediction model, and outputting an electric energy substitution potential prediction result. According to the method, the parameters of the sparrow search algorithm are configured, the chicken swarm optimization algorithm and the Cauchy-Gaussian mutation strategy are combined, global and local search optimization of prediction model parameters is realized, the search efficiency of the algorithm is improved, the ability of the algorithm to jump out of a local optimal solution is enhanced, and therefore the generalization ability and the prediction precision of the prediction model are improved.

The invention discloses an unmanned aerial vehicle path-finding method and system based on a multi-strategy improved artificial travel mouse algorithm, and the method comprises the steps: collecting the current three-dimensional coordinates of an unmanned aerial vehicle, and setting a starting point and an ending point of the flight of the unmanned aerial vehicle; constructing a target function by taking the minimum height, the corner cost and the shortest flight distance as targets; substituting the current three-dimensional coordinate of the unmanned aerial vehicle into the target function to evaluate the fitness, introducing improved Chebyshev mapping, Gaussian distribution and a Cauchy Gaussian mutation strategy to improve an artificial travel mouse algorithm, and obtaining the optimal next three-dimensional coordinate of the unmanned aerial vehicle by using the improved artificial travel mouse algorithm; and forming an optimal operation path of the unmanned aerial vehicle according to the optimal three-dimensional coordinates, and adjusting the direction of the unmanned aerial vehicle based on the optimal path. According to the method, the flight path of the unmanned aerial vehicle is optimized by improving the artificial travel mouse algorithm, so that the search efficiency of the unmanned aerial vehicle is improved, the navigation cost is saved, and the search requirements of complex environments and limited cost are met.

The embodiment of the invention provides a thermal power station valve inner leakage intelligent identification method based on an improved SVM. The method comprises the steps that a training sample set is constructed by adopting data corresponding to parameters during operation of a drain valve of a thermal power station; optimizing the Kepler optimization algorithm KOA based on the Sobol sequence and the adaptive Gaussian mutation to obtain an improved Kepler optimization algorithm IKOA; the SVM model is trained through a training sample set by adopting multi-step sliding, key parameters of the SVM model are self-optimized by adopting an improved Kepler optimization algorithm IKOA, and the SVM model constructed by adopting optimal values of the key parameters is used as an intelligent identification IKOA-SVM model of the valve inner leakage amount. A Sobol sequence is adopted to initialize a KOA population to achieve optimization of local search and global search, an optimized IKOA-SVM model is obtained, and intelligent identification of the inner leakage of the thermal power station valve is achieved.

This invention discloses a precise motion controller for a magnetically controlled capsule robot based on IDMO-PID, comprising a magnetically controlled capsule robot and an external magnetic source. The magnetically controlled capsule robot is placed in the stomach of a human, and the external magnetic source is located on the outside of the human body. This invention optimizes the motion control system of the capsule robot by rationally analyzing the motion mode and spatial magnetic field of the capsule robot. Based on the linearly decreasing inertia weight factor and the Cauchy-Gaussian mutation strategy, the population position update stage and the optimal solution output stage of the dwarf meerkat optimization algorithm are optimized, which effectively improves and reasonably balances the global search capability and local search capability of the basic dwarf meerkat optimization algorithm at different stages, thereby improving the convergence progress and convergence speed of the algorithm. The improved dwarf meerkat optimization algorithm is used to optimize the motion control system parameters of the capsule robot, realizing the precise motion of the capsule robot under multi-source spatial magnetic fields.

The invention discloses a power distribution network fault recovery method based on a distributed energy grid-connected point connectivity constraint. Firstly, a pre-reconstruction scheme set is generated as a primary particle swarm; calculating system power flow and node voltage of each pre-reconstruction scheme under double time sections at the initial stage and the later stage of recovery; then, taking the maximum weighted load recovery amount and the minimum switching action times as target functions of power distribution network fault recovery control, introducing a penalty factor, constructing a fitness function, and calculating the fitness of each particle; and finally, when the number of iterations reaches a set number of times, performing dominant Gaussian mutation operation on the particle swarm, namely randomly overturning partial switch states in a load transfer loop, and not disconnecting adjacent switches of the distributed new energy grid-connected nodes which are in a connected state any more. And updating the speed and the position of each particle, and taking the pre-reconstruction scheme corresponding to the globally optimal particle as the optimal scheme. The method improves the fault recovery speed and the power supply reliability, and avoids the long-time power failure of the load side.

The application belongs to the technical field of robot motion control, and discloses a vehicle control parameter setting method based on a hybrid multi-strategy melolontha optimization algorithm. The method designs a hybrid multi-strategy melolontha optimization algorithm based on a traditional melolontha optimization algorithm, and proposes an optimization algorithm for population initialization and population position updating. A good point set theory is introduced to uniformly generate the population position in the initialization stage, so as to improve the diversity of the initialized population. After the position updating, the global search ability and the convergence speed of the algorithm are improved by adding the gravity reverse learning search and the Gaussian mutation in each dimension. The method is suitable for parameter setting optimization of an ADRC controller. The method not only can improve the diversity of the position updating mode of an amphibious vehicle, but also can improve the global search ability and the convergence speed of the algorithm while avoiding the algorithm from falling into a local optimal solution, so as to further improve the position and attitude control performance of the amphibious vehicle.

The invention discloses a network construction type SVG parameter identification method based on an adaptive fireworks algorithm, and the method comprises the steps: constructing a joint variable weight sensitivity index, and determining a to-be-identified parameter through sensitivity analysis; constructing a mapping relation by adopting ConvLSTM based on a hierarchical attention mechanism to obtain response waveforms under different parameter combinations; and identifying the control parameters and the electrical parameters of the networking type SVG based on an adaptive fireworks algorithm to obtain real-time accurate parameters of the networking type SVG. According to the method, the adaptive fireworks algorithm is adopted to perform parameter identification of the network-forming SVG, multi-dimensional all-dimensional optimization is processed through Gaussian variation and mapping, a globally optimal solution can be quickly converged, automatic adjustment is performed in the aspects of explosion intensity, explosion amplitude and the like, and adaptive optimization is realized. According to the method, the objectivity of selection of the to-be-identified parameters and the parameter identification efficiency are effectively improved, the real-time accurate parameters of the network construction type SVG can be obtained, and data support is provided for stable operation and optimization control of the network construction type SVG.

The invention relates to the technical field of swarm intelligence, in particular to an infrastructure site dynamic job scheduling method based on multi-agent collaboration, and the method comprises the steps: constructing a task state matrix; in optimization algorithm iteration, the positions of population individuals are decoded into task priorities, and the predicted execution time of each task node is obtained in combination with a task state matrix; calculating a process coupling degree used for representing a process connection tightness degree; counting the process coupling degree of the population individuals in the current iteration, calculating the evolutionary stagnation risk in combination with the standard deviation of the process coupling degree in the initial iteration, and selecting a target strategy from the Cauchy variation strategy and the Gaussian variation strategy to update the population individuals; and converting the global optimal position at the end of iteration into a scheduling instruction set. According to the technical scheme, the invalid waiting time between procedures can be shortened, and the comprehensive operation efficiency of an infrastructure site is improved.

The invention discloses a path planning method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining environment data through a sensor, and constructing a static environment model according to the environment data; determining an initial position of a target object in the static environment model; using a multi-strategy improved dung beetle algorithm to determine an initial path corresponding to the initial position in a static environment model; the initial path is evaluated to obtain an evaluation result, the evaluation result is utilized to update the initial path to obtain a target path, the evaluation result of the target path is that the evaluation reaches the standard, efficient path planning of the target object in the complex static environment is achieved, and through combination of a golden sine algorithm and a Cauchy Gaussian mutation algorithm, the path planning efficiency of the target object in the complex static environment is improved. The convergence speed of path planning is effectively improved, and the working efficiency is greatly improved.

The present invention discloses a method and system for optimizing energy storage configuration in an active distribution network based on an improved Parrot algorithm. The method includes: constructing a multi-objective energy storage optimization configuration model based on preset constraints, taking node voltage offset, energy storage cost, and distribution network loss as objective functions; introducing chaos theory to initialize the Parrot algorithm population; dynamically adjusting the weights of foraging behavior, staying behavior, communication behavior, and fear of strangers in the Parrot algorithm during each iteration based on the individual fitness value and the current iteration number; selecting Gaussian mutation or Cauchy mutation based on the mutation probability, performing mutation operations on the Parrot individuals during the iteration process, and ultimately obtaining an improved Parrot algorithm; solving the multi-objective energy storage optimization configuration model based on the improved Parrot algorithm, and outputting the energy storage optimization configuration results. This method achieves efficient collaborative optimization of multiple objective functions, providing more accurate and comprehensive decision support for the energy storage optimization configuration of active distribution networks.

The invention discloses a robot path planning method based on a self-adaptive coevolution locust optimization algorithm, relates to the technical field of mobile robots, and aims to solve the problems that an existing GOA algorithm is not reasonable in parameter calculation mode, prone to falling into local optimum and insufficient in overall search capability. According to the robot path planning method based on the adaptive coevolution locust optimization algorithm, a cosine adaptive strategy is introduced, a calculation mode of a regulation coefficient c in GOA is improved, the balance of global and local search capabilities of the algorithm is well realized, a cooperative strategy of initial sub-population and inter-population information sharing is adopted, and the algorithm can be applied to robot path planning. According to the method, the influence of random initialization on the algorithm is reduced, the condition of individual information redundancy in the initialized population is improved, Gaussian mutation operation is performed on the population after the population is updated every time, and the probability of finding the optimal individual is improved while the number of inferior solutions in the population is reduced.