32 results about "Genetic programming" patented technology

The invention discloses an integrated circuit fault supervision method and system for dynamic time slot allocation, and the method comprises the steps: collecting and preprocessing multi-source time slot data, and generating a standardized data set; based on the standardized data set, genetic programming is used for generating a mathematical formula for describing the relation between time slot allocation and fault detection; clustering the standardized data set by using fuzzy C-means clustering, and identifying a data state category; dynamically adjusting a time slot allocation strategy in combination with a mathematical formula and a clustering result, and generating a repair result; if abnormity is detected, an adjustment formula is quickly generated based on a mathematical formula and a clustering category, and repair is executed; and feeding back the distribution and repair result, and adaptively updating the mathematical formula and the clustering result. According to the invention, efficient adaptive optimization of dynamic time slot allocation and intelligent fault supervision in an integrated circuit system is realized.

The invention provides a complex constraint-oriented time parallel unmanned aerial vehicle cluster dynamic planning method based on adaptive genetic programming, and belongs to the field of unmanned aerial vehicle path planning, and the method comprises the steps: initializing a genetic programming expression population; constructing a time parallel unmanned aerial vehicle cluster path planning model and performing adaptive value evaluation on the output of each individual in the population; the adaptive value loss of the failure scheme is set as dynamic penalty of self-adaptive change; and performing iterative optimization on the population based on survival of the fittest, and applying the optimal individual as a final path scheme to the unmanned aerial vehicle cluster to execute the inspection task. According to the invention, after the unmanned aerial vehicle takes off from the specified charging station, all inspection tasks can be completed in parallel under limited electric quantity, electric quantity supplementation is carried out through the charging station, and the unmanned aerial vehicle finally returns to the original station for maintenance and charging; according to the method, under the condition of complex problem constraint, approximate optimal path node selection can be effectively generated, a reasonable task allocation strategy and a routing inspection route are obtained, and the task completion efficiency is remarkably improved.

The invention discloses a hyper-heuristic order grouping and robot sorting path integrated planning method based on genetic programming. The method comprises the following steps: acquiring order demand information and warehouse layout information of an intelligent manufacturing workshop production system; determining an objective function of order grouping and sorting path planning; establishing a double-commodity network flow model integrating order grouping and path planning; setting constraint conditions for the double-commodity network flow model; a priority function rule is generated in a genetic programming hyper-heuristic mode, a dynamic programming strategy is combined, the double-commodity network flow model is solved, and the optimal result of order grouping and batching and robot sorting paths is obtained. According to the method, the sorting efficiency optimization problem of multi-variety and high-frequency orders in the intelligent manufacturing feeding warehouse is solved, and the solving efficiency and rule generalization ability can be remarkably improved, so that the sorting time is effectively shortened, the intelligent manufacturing short window requirement is met, and theoretical guidance and practical support are provided for improvement of the warehousing feeding efficiency.

The invention relates to the technical field of industrial scheduling, and discloses a flexible job shop self-organizing scheduling-oriented multi-target dual hyper-heuristic method, which comprises the following steps of: generating a process selection rule and an interval selection rule by utilizing a genetic programming rule, and generating a rule set; based on a deep reinforcement learning method, a dynamic decision strategy is constructed in combination with a rule set, and multi-step action sequence optimization is carried out when self-organizing scheduling is triggered; and generating an instance through random combination of parameter indexes of a predetermined dynamic event, and calculating a performance index according to the generated instance to realize feasibility verification. Through a dynamic collaborative optimization mechanism of self-organizing scheduling, an autonomous decision closed loop can be realized under multiple disturbances such as equipment failure, processing fluctuation, new order insertion and the like, and the self-healing capability, response speed and anti-interference toughness of a production system in a dynamic environment are remarkably improved; and an intelligent decision-making scheme with an autonomous evolution capability is provided for a complex scheduling problem in an intelligent manufacturing scene.

The invention belongs to the technical field of intelligent manufacturing, and discloses a digital twinning workshop real-time scheduling method for limited transportation resources and charging constraint scenes, which comprises the following steps: constructing a real-time scheduling framework based on digital twinning and deep reinforcement learning to realize real-time interaction of virtual and real data; proposing a two-stage real-time scheduling model based on deep reinforcement learning; taking the minimum completion time as a target, and establishing a Markov decision process of the DFJSP-LTR-C; five key elements including an interaction point, a real-time scheduling process, a real-time state feature, an action space based on improved genetic programming and a composite reward function are designed for the scheduling model; an IAD3QN training method based on a multi-head attention mechanism is provided to train a scheduling agent. According to the method, adaptive collaborative optimization of production and transportation resources is realized under the condition of considering limited transportation resources and charging constraints, the maximum completion time is effectively reduced, and the feasibility and robustness of a scheduling scheme in an actual workshop are improved.

Method and apparatus (10) for controlling a modification process of hygroscopic material (15) comprising method steps of: measuring at least one process variable from the modification process at least during the modification; measuring at least one process variable from the hygroscopic material at least during the modification; calculating at least one intermediate control parameter by a neural network busing at least the measured process variables as input parameters of the neural network; and controlling the modification process by using genetic algorithms and genetic programming based on the said at least one intermediate control parameter determined by the neural network.

The invention discloses a dual-tree genetic programming method for processing a long-tail image classification problem, and the method comprises the following steps: constructing a dual-tree genetic programming individual which comprises a data enhancement tree and a feature extraction tree; setting a staged sampler; performing a multi-generation evolution process, wherein each generation comprises population initialization, individual fitness evaluation, selection, crossover and mutation operation; after the evolution is finished, outputting a dual-tree individual with the highest fitness as a final classification model for classifying a new image; the method is simple and efficient in model deployment and has practical value.

A method for detecting changes in remote sensing images based on genetic programming is disclosed. This method is used to detect changes in two remote sensing images of the same area taken at different times. The method includes the following steps: Step A: Two remote sensing images of the same area taken at different times are acquired using a computer, and the two images are calibrated using the same coordinate system. Step B: From the two remote sensing images selected in Step A, at least 100 areal samples are selected as sample data. Step C: Based on the sample data obtained in Step B, a threshold function for comparing the two remote sensing images is calculated using genetic programming to detect changes between the two images. The genetic programming-based method for detecting changes in remote sensing images provided in this application can fully explore the implicit information between the features of images from different time periods, effectively improving the accuracy of change detection and significantly increasing detection efficiency.

This invention relates to the fields of artificial intelligence and computer vision, and provides a few-shot fine-grained image classification method based on multi-scale feature extraction, attention, and relation enhancement using genetic programming. The method includes: extracting multi-scale deep features using a pre-trained deep feature extraction network (preferably ResNet-18); designing a strongly typed genetic programming tree structure comprising an input layer, a multi-scale feature extraction layer, a channel selection layer, an attention layer, a feature concatenation layer, and an output layer; enhancing discriminative regions and suppressing background noise through a spatial attention mechanism; developing a prototype-query relation enhancement module to amplify the regions of common interest between the support set and the query set to enhance feature representation; and using the sum of classification accuracy and cluster separation score as a fitness function to guide the evolutionary search. This invention combines deep neural network representation with genetic programming evolutionary search, achieving excellent classification performance on fine-grained few-shot image classification tasks while maintaining interpretability.

The invention discloses a method for generating circuit topology based on a transfer function, relates to the technical field of circuit design, and solves the technical problem that it is difficult to combine an artificial intelligence method in circuit design to improve the efficiency and reliability of analog circuit design. The method comprises the following steps: based on a graph theory, establishing a connected graph G under a given RLC circuit structure and element parameters; deriving the relationship between the transfer function and the impedance or admittance of the network nodes of the connected graph G based on the RLC circuit; searching and calculating a tree and a secondary tree required by the transfer function, and calculating to obtain a final expression of the transfer function; analyzing the order and the structure of the transfer function, and combining specification constraints and design specifications of circuit design to obtain a circuit design rule; and searching topology and parameters of the RLC circuit through genetic programming to obtain an optimal topological structure and an optimal element admittance value. According to the method, automatic search of genetic programming is guided by analyzing the hard constraint rule set, and the efficiency and reliability of circuit comprehensive design are remarkably improved.

The invention discloses an eVTOL collaborative design optimization method based on improved genetic programming, and the method comprises the steps: firstly constructing pneumatic, propulsion and dynamics high-fidelity models of an eVTOL takeoff stage, building an original collaborative optimization problem, and solving the problem through a direct transcription method; secondly, multi-target genetic programming fusing a root mean square error, an akaike information criterion, a predicted residual sum of squares and a Bayesian information criterion is creatively adopted, and an agent model for calculating thrust is constructed to replace a high-fidelity thrust model to solve and obtain optimal design parameters. The error between the optimal solution obtained by the method and the optimal solution obtained by using the high-fidelity model is very small, the calculation cost is remarkably reduced, and multidisciplinary efficient collaborative optimization is realized.

The invention discloses a pre-training model guided genetic programming-based battery life prediction method. The method comprises the steps of defining a battery life prediction target and battery cycle characteristics; the method comprises the following steps: generating a data point pair set by using a mathematical expression containing battery cycle characteristics, constructing a training data set, and pre-training a Transform model; collecting battery data, and constructing a battery data set; inputting the battery data set into a Transform model, generating a guide sub-population by using the Transform model, and combining the guide sub-population with the generated random sub-population to construct a hybrid initial population of genetic programming; iterative evolution is carried out on the mixed initial population, mutation operation is carried out, and the mutation operation is executed based on semantic back propagation and semantic guidance of a Transform model; and judging whether the evolution reaches a termination condition, if the evolution is terminated, outputting a mathematical expression, and inputting the battery cycle characteristics of the to-be-tested battery into the mathematical expression to obtain a battery life prediction result.

The application discloses a data efficient image classification method based on multi-tree genetic programming, comprising the following steps: obtaining an image to be classified; pre-processing the image to be classified; a genetic programming individual respectively processes red, green and blue channels of the image, and performs multi-scale operation, region extraction, feature extraction and feature connection on the channels; the individual obtained through evolution is used for feature conversion on the image to be classified to obtain new features and normalization; and the new features are input into a classifier to output a classification result of the image to be classified. The color feature extraction method based on the multi-tree genetic programming can learn color features with discrimination and rich information, thereby improving classification accuracy; and the multi-scale feature extraction can enhance the ability of capturing more comprehensive and rich image features for classification.

The application belongs to the field of artificial intelligence and finance, and relates to a risk identification factor mining method, which comprises the following steps: obtaining different data types of data, setting different types of operators according to an insurance risk identification target, constructing a genetic programming model according to the different types of operators, initializing a formula population, iteratively calculating a risk explanation degree, continuously optimizing a formula tree through a genetic algorithm, obtaining a final formula tree population, selecting a formula tree with the highest risk explanation degree from the final formula tree population as a risk identification factor, and testing and optimizing the risk identification factor to obtain an optimal risk identification model. The application also provides a risk identification factor mining device, a computer device and a storage medium. In addition, the application also relates to blockchain technology, and text data and numerical data can be stored in the blockchain. The application can better realize automatic identification of insurance risks and improve the accuracy of risk identification.

The application discloses a photovoltaic power station power prediction method based on prior knowledge guided genetic programming, first, obtaining training data preprocessed by symbol normalization, learning photovoltaic power generation symbol rules through self-supervised pre-training of a Transformer model, and outputting a prior symbol vector of a target function; a double fitness collaborative optimization mechanism containing numerical fitting accuracy and physical mechanism similarity is designed, and the prior vector is used to strengthen and select individuals with high physical matching degree; a Pareto multi-objective optimization is used to balance the double-dimensional conflict, the prior knowledge is transferred to the whole evolution process of genetic programming, and finally the optimal explainable prediction model is extracted from the Pareto frontier solution set. The application solves the problems of traditional genetic programming evolution blindness, easy falling into local optimum and poor physical consistency of generated model, realizes explicit physical explainability of the model while ensuring prediction accuracy, and can be widely applied to power generation scheduling and operation optimization of photovoltaic power stations.

A method is provided that includes accessing a multivariate time series of flight data for an aircraft, and iteratively performing runs of genetic programming on groups of the sensors. A population of computer programs is randomly generated from a selected group of the plurality of sensors, and primitive functions selected from a library of primitive functions. The population is iteratively transformed into new generations of the population, and includes sub-rankings of the group of sensors based on a quantitative fitness determined according to selected fitness criterion. A ranking of the group of sensors from the sub-rankings of the group of sensors is produced. An aggregate ranking of the plurality of sensors is produced from the ranking of the group of sensors over a plurality of iterations. And the subset of sensors is selected from the aggregate ranking of the plurality of sensors, and according to selected optimization criterion.

The invention relates to the field of traditional Chinese medicine diagnosis, and discloses a traditional Chinese medicine diagnosis and treatment knowledge graph construction method and system based on knowledge reasoning, and the method comprises the steps: constructing a traditional Chinese medicine diagnosis and treatment knowledge graph, and carrying out the node attribute distribution and node relation determination; on the basis of a mode of combining anomaly detection of statistical analysis and machine learning, abnormal nodes in the traditional Chinese medicine diagnosis and treatment knowledge graph are judged, and root cause node alternative items are identified; the centrality of each node in the traditional Chinese medicine diagnosis and treatment knowledge graph is analyzed and calculated through a graph theory, and root cause node alternative items are preliminarily sorted according to the centrality; optimizing the sorting rule of the root cause node alternative items by using genetic programming and a causal discovery algorithm to obtain sorted root cause node alternative items, and determining root cause nodes; and constructing a diagnosis and treatment path knowledge graph according to the root cause nodes, the related diagnosis and treatment information nodes and the causal relationship among the nodes. According to the invention, the accuracy and efficiency of the traditional Chinese medicine diagnosis and treatment path are improved, and the root cause node identification sorting is optimized.

The invention discloses a battery capacity degradation curve prediction model based on a genetic programming optimization encoder. The method comprises the following steps: firstly, acquiring time sequence data of battery charging and discharging multiple signals, and constructing a variational auto-encoder for each signal; replacing an encoder of the auto-encoder with a genetic programming GP structure, and capturing a complex nonlinear relationship between signals; and a multi-tree GP structure is adopted and pruning optimization is carried out, so that the model precision and efficiency are improved. And combining the output of the optimized GP encoder into a strong encoder, and carrying out joint modeling on multiple signals. And an artificial neural network (ANN) is used as a decoder to decode the low-dimensional features generated by the GP encoder to predict the battery capacity. And training an ANN decoder in combination with the real capacity data to generate a battery capacity decline curve. The method can accurately capture the state change rule of the battery, and has wide application prospects in the fields of battery management, life prediction, health monitoring and the like.

This invention discloses a method and system for constructing a rainfall-runoff model. The method involves collecting and preprocessing relevant daily sequence data within a watershed; constructing a hydrological model component library; pre-assembling the hydrological model structure; defining a parameter space; performing a collaborative search using genetic programming; simulating daily runoff using the optimal model structure and parameters obtained through the search and calibration; evaluating the model's generalization performance; and outputting the optimal model structure type, optimized parameter vector, calibration set, validation set performance indicators, and runoff simulation sequence. This invention can automatically generate rainfall-runoff models with adaptive model structures and interpretable hydrophysical mechanisms, effectively improving the automatic discovery capability of model structures and the interpretability of hydrophysical mechanisms, while reducing manual workload while avoiding overfitting.

This invention discloses a modeling method and system for a hydrogen-powered drone swarm refueling scheme. The method involves: decomposing the drone's flight process into multiple independent events; managing these independent events using a three-alignment principle; calculating the energy consumption of the drones during flight based on a nonlinear energy consumption model; rasterizing the continuous flight space; planning obstacle avoidance for the mission drones' flight paths; and using stacked genetic programming to cyclically insert random points as refueling rendezvous points into the mission drones' flight paths, assigning them to the corresponding refueling drone paths to form a refueling scheduling scheme. This invention can effectively model complex constraints such as time, space, and energy in a scenario, providing a detailed model and planning framework for collaborative refueling tasks of drone swarms. This reduces the difficulty of subsequent design and optimization algorithms, improves solution efficiency and feasibility, and is expected to enhance the accuracy of drone planning.

The application discloses a method for generating a circuit topology based on a transfer function, and relates to the technical field of circuit design, and solves the technical problem that it is difficult to combine artificial intelligence methods to improve the efficiency and reliability of analog circuit design in circuit design. The method comprises the following steps: based on graph theory, a connected graph G is established under the structure and element parameters of a given RLC circuit; the relationship between the transfer function and the impedance or admittance of the network nodes of the connected graph G is derived based on the RLC circuit; a tree and a secondary tree required for searching and calculating the transfer function are searched and calculated to obtain the final expression of the transfer function; the order and structure of the transfer function are analyzed, and circuit design rules are obtained in combination with specification constraints and design specifications of the circuit design; and the topology and parameters of the RLC circuit are searched through genetic programming to obtain the optimal topology structure and the optimal element admittance value. The application significantly improves the efficiency and reliability of circuit synthesis design by analyzing a set of hard constraint rules to guide the automatic search of genetic programming.

The application discloses an unmanned aerial vehicle (UAV) inspection path planning method based on automatic target differential evolution, which comprises the following steps: standardizing modeling of a UAV inspection area, setting a plurality of intermediate control points between adjacent inspection points, encoding the UAV inspection path, and constructing a fitness function; generating an initial seed solution according to the spatial distribution characteristics of the inspection points, and generating an initial population by using the initial seed solution; one individual in the initial population represents a UAV inspection path; a target vector generation operator is obtained through genetic programming offline training; the initial population is updated by target differential evolution according to the target vector generation operator to obtain a test individual; the test individual is repaired, and the optimal individual is selected according to the fitness function to enter the next generation population iteration; and the planning result is output when the termination condition is reached. The method improves the search ability and convergence efficiency of the high-dimensional path coding problem, and greatly optimizes the UAV inspection path.

The application discloses a kind of multi-objective synthetic oversampling methods for unbalanced data, based on genetic programming, the synthesis of minority class sample is modeled as constraint multi-objective optimization problem, including: genetic programming driven batch synthesis: individual is coded as a complete sample synthesis rule, a batch of high-quality samples is automatically generated in single run by evolution search, capture complex nonlinear relationship beyond linear interpolation;Constraint multi-objective evaluation mechanism: the sample generation process is modeled as constraint multi-objective optimization problem, while optimizing inter-class separation and local neighborhood structure, and from classification boundary, distribution consistency, sample diversity, sample quality four aspects explicitly constraint sample, ensure the effectiveness of generated sample;Constraint processing mechanism: design special constraint processing strategy to select feasible solution to participate in sample generation, promote population feasibility and realize batch synthesis of high-quality minority class sample;The application significantly improves sample generation efficiency, is advantageous to process large-scale unbalanced data.

This invention discloses a battery life prediction method based on pre-trained model-guided genetic programming, comprising: defining a battery life prediction target and battery cycling characteristics; generating a set of data point pairs using a mathematical expression containing battery cycling characteristics, constructing a training dataset, and pre-training a Transformer model; collecting battery data and constructing a battery dataset; inputting the battery dataset into the Transformer model, using the Transformer model to generate a guide subpopulation, and combining it with a randomly generated subpopulation to construct a hybrid initial population for genetic programming; iteratively evolving the hybrid initial population and performing mutation operations, the mutation operations being executed based on semantic backpropagation and semantic guidance of the Transformer model; determining whether the evolution has reached the termination condition, if the evolution terminates, outputting a mathematical expression, inputting the battery cycling characteristics of the battery to be tested into the mathematical expression, and obtaining the battery life prediction result.

The application relates to a method for solving a complex product assembly sequence planning problem considering human factors, wherein the method comprises: comprehensively introducing cognitive, physiological, psychological, skill and organizational human factors in the complex product assembly sequence planning process, and optimizing and solving the assembly sequence based on genetic programming. By effectively deciding the human factor data acquisition mode, the state data of workers are acquired, the assembly sequence is adjusted in combination with the constraint of the human factor, and finally the optimal assembly sequence planning considering the state of workers is realized, so that the efficiency and reliability of the assembly process are improved. Therefore, the problems that the prior art relies on the experience of engineers for manual planning, human factors are difficult to be comprehensively considered, and modern complex product manufacturing requirements cannot be met are solved.

The disclosure concerns methods and systems for controlling charging processes for charging electric vehicles by a charging system based on charging control policies. The disclosure provides approaches for automated generating of charging control policies. The system acquires historical information on charging parameters and battery parameters, and determines, for each time step, information on the available total amount of energy for charging the electric vehicles. The system computes, for each time step, and for each charging control policy of a plurality of charging control policies, a fraction of the total amount of energy for charging the electric vehicles with the charging control policy. The system controls charging for each time step based on the plurality of charging control policies and the computed fraction of the total amount of energy for each charging control policy. The disclosure further proposes an automated generating of charging control policies using a genetic programming approach.

The present invention provides a genetic programming hyper-heuristic traffic flow allocation method based on knowledge transfer, comprising the following steps: obtaining road network information; abstracting the road network into a weighted directed graph; constructing a traffic simulation model based on the road network information and determining an optimization target; evaluating each routing strategy in the population using the traffic simulation model; extracting knowledge based on the source domain; and when the problem scenario changes, iteratively optimizing the target domain based on the knowledge learned from the source domain, and applying the optimal individual as the final routing strategy to traffic flow allocation. The present invention effectively solves the problem of the effectiveness of routing strategies for traffic flow allocation when the road network structure changes, while improving the efficiency of retraining in new scenarios. By applying transfer learning to traffic flow optimization, the general knowledge learned in the source domain is transferred to the target domain, which can significantly improve the efficiency of retraining under the new road network structure, is applicable to traffic flow allocation strategies of different types of road network structures, and improves the generalization ability of routing strategies.

This invention discloses a genetic programming mutation probability optimization method based on the maximum mutual information coefficient. In traditional genetic programming, feature selection tends to shift from initial completely random selection to selective selection with bias. This invention constrains the search direction of genetic programming, thereby improving search efficiency. The method includes: Step S1, using the maximum mutual information coefficient to measure the correlation between each feature and the target in the dataset, and merging the correlations of each feature and the target into a correlation vector; Step S2, determining the probability distribution of genetic programming when selecting new features through mutation based on the correlation vector; Step S3, performing genetic programming evolution, during which the probability distribution remains fixed and is unaffected by the number of features selected within the population. This invention reduces the impact of random initialization on the overall performance of genetic programming, enhances the robustness of genetic programming on fundamental problems, improves training efficiency and accuracy, and enhances model generalization performance.

The application relates to a dynamic air traffic flow regulation method, and belongs to the technical field of air traffic flow management. The method solves the problems that global airspace information cannot be utilized, airspace resources cannot be fully utilized, and a decision is not comprehensive in the prior art. The method comprises the following steps: S1, obtaining initial data and building a simulation environment according to the obtained initial data; S2, determining a state set and a function set according to the initial data, which are used to generate multiple-tree individuals; S3, setting selection operators, crossover operators and mutation operators of the multiple-tree individuals in genetic programming; S4, setting a multiple-tree individual initialization method to obtain multiple-tree individuals with a set population number; S5, calculating fitness values of the multiple-tree individuals based on the simulation environment; S6, performing evolution learning based on the obtained multiple-tree individuals, and outputting a trained best multiple-tree individual and an optimal rule; and S7, simulating the obtained best multiple-tree individual and the optimal rule to obtain and output a scheduling scheme.

A method and apparatus (10) for controlling a reforming process of a hygroscopic material (15) includes the steps of measuring at least one process variable from the reforming process at least during the reforming, measuring at least one process variable from the hygroscopic material at least during the reforming, calculating at least one intermediate control parameter by a neural network by using at least the measured process variable as an input parameter for the neural network, and controlling the reforming process using a genetic algorithm and genetic programming based on the at least one intermediate control parameter determined by the neural network.