58 results about "Evolutionary computation" patented technology

The invention discloses a path planning heuristic function generation platform and method based on collaborative optimization of a large-scale language model and evolutionary computation. According to the technology, the large-scale language model (LLM) and evolutionary computation (EC) work cooperatively. The platform generates or mutates a heuristic function expressed as an executable code through LLM based on a structured prompt containing an environment context and performance feedback; and an EC framework (such as genetic programming) is combined with performance evaluation feedback to perform selection and iterative optimization on a heuristic code population, and population diversity is maintained. The method aims at overcoming the limitation that a traditional heuristic design is difficult and poor in adaptability, a high-quality heuristic function adapting to a complex and dynamic environment is automatically generated, and therefore the efficiency of a path planning algorithm and path quality are remarkably improved.

The invention relates to the technical field of data processing, in particular to a multi-modal sensor data real-time fusion processing method for a smart community. According to the method, a sensor network topological graph is constructed, a connection weight is optimized, distributed clock synchronization is realized by using a graph Laplacian matrix, and clock drift prediction and compensation are performed in combination with a fractional Brownian motion model; performing wavelet transform decomposition on the sensor data after time sequence alignment, calculating each scale Hurst index, predicting a load trend through a fractal prediction model, and outputting an optimal resource allocation scheme through hybrid evolution calculation; the method comprises the following steps: constructing multi-modal sensor data into a graph structure, extracting node features by using a graph convolutional neural network, obtaining global feature representation by using a self-attention mechanism, and performing anomaly detection classification in combination with a resource utilization rate and a prediction error; an anomaly detection feedback mechanism is established, and Laplacian matrix eigenvalues and weight parameters are dynamically adjusted; the real-time performance, the accuracy and the robustness of data fusion processing are improved.

The invention discloses a foot type gait and form co-evolution system integrating reinforcement learning and evolutionary computation, and relates to the technical field of robots and the field of artificial intelligence. Comprising the steps that a real-time motion control strategy learning module is realized based on a dynamic mask integrated double-Q learning method; the gait and form collaborative optimization module optimizes key gait and form parameters influencing the global motion performance of the foot robot based on an evolutionary computation theory; the training environment adaptive enhancement module comprises an automatic domain randomization module and a course learning module; the parameterized trajectory generator module is used for providing structured motion prior for the real-time motion control strategy learning module; the robot simulation environment is used for simulating the physical effect of the robot and the environment in the computer; according to the invention, the high-speed, stable and efficient autonomous movement capability of the foot robot in a complex and changeable unstructured environment is realized.

The present invention introduces an intelligent coal gangue washing method guided by deep reinforcement learning and evolutionary computation. It involves several steps: S1 involves installing various sensors at key control points of a jig to achieve comprehensive, real-time data acquisition and maintain consistent data collection frequencies; S2 includes gathering data on a server via OPC protocol, using deep reinforcement learning to devise control strategies for jig operations under good communication, and employing evolutionary algorithms when communication is disrupted; S3 entails sending these control strategies back to the control unit through OPC protocol, enabling automated operation of the jig. This method enhances jig operation efficiency through intelligent control, utilizing deep learning, evolutionary computation, and surrogate models to optimize performance even when operational data is incomplete.

The invention discloses an AI business process automatic generation method based on AI model combinatorial optimization, and the method comprises the following steps: constructing a multi-level AI capability decoupling and reconstruction module, and carrying out the bottom-up hierarchical modeling and top-down modular decoupling, dynamic mapping of an AI atomic power layer, an AI modular production capacity layer, an AI general capability layer and an AI application business layer is realized; an elastic AI capability combinatorial optimization module is constructed, AI capability combinatorial optimization oriented to three dimensions of data, features and models is carried out based on service quality requirements, and multiplexing, combination and arrangement of AI capabilities are realized through evolutionary computation optimization driven by an agent model; and an AI business process automatic generation module is constructed, an AI service containerization deployment scheme is generated and optimized based on data-driven process mining and a hyper-heuristic algorithm, and AI business process automatic generation is realized. According to the invention, the adaptability and execution efficiency of the AI technology in a complex scene can be improved.

The invention discloses an automatic algorithm configuration method based on deep reinforcement learning in evolutionary computation. The method comprises the following steps: determining a problem set; obtaining initial data for constructing a Markov decision process; defining a state vector of each individual in a time step in an evolutionary computation algorithm by using an individual state vector according to the initial data; constructing feature vectors of population features, individual features and development features according to the initial data; defining a continuous action space according to the initial data, and jointly controlling the selection of all individuals on hyper-parameters; based on whether a better solution is found or not, a reward mechanism is set, and maximization of expected benefits is achieved; and the deep reinforcement learning agent uses an attention mechanism to control dynamic hyper-parameters of the algorithm, so that a Markov decision process is realized, and automatic algorithm configuration is completed. According to the method, direct mapping from high-dimensional sensing information to continuous action space output is realized, configuration of individual exploration and development tradeoff is dynamically adjusted, and a balance mechanism of exploration and development is optimized.

An analogy of Post-Selections is: “Somebody claims that his scheme provided a lottery ticket number that has won $1M, but he conceals that the scheme has spent 2 millions of lottery tickets of $1 each. The reported ticket is only the luckiest. The luckiest ticket was Post-Selected after the actual lottery test. The luckiest lottery ticket will not have the same luck next time.” Many machine learning methods suffer from Post-Selections, from neural networks, to reservoir computation, to swam intelligence to evolutionary computation. The numbers $1M, 2M and $1 and the chance to win in the analogy differ across different machine learning problems, but the nature of the flaw in the reports is basically the same. This invention presents a method that does not need any Post-Selections since it trains only one network that is computed in a closed form that corresponds to the most-probable network from training experience.

The invention discloses an abnormal sample detection and identification method and system based on evolutionary computation and multi-modal consistency constraint, and belongs to the field of network security and artificial intelligence security. The method comprises the following steps: S1, acquiring and preprocessing input data; s2, multi-modal feature extraction and unified characterization are carried out; s3, optimization feature selection and weight self-adaption are carried out; s4, constructing a cross-modal fusion detection model; s5, abnormal sample risk identification; and S6, result judgment output: generating a final detection label according to a threshold function. The method shows high accuracy and strong robustness in counterfeit image and video detection and abnormal sample elimination, and can be widely applied to the fields of multimedia authentic identification, intelligent security and protection and AI content traceability.

The invention provides a drug micromolecule multi-attribute optimization method combining deep learning and evolutionary computation, which comprises the following steps: S1, modeling a drug micromolecule multi-attribute optimization problem into a multi-objective optimization problem, and screening drug micromolecules meeting required characteristics from a DugBank database to obtain an initial molecular population; s2, fragmenting each drug small molecule of the molecular population, embedding the drug small molecule into a continuous submerged space, and carrying out molecular coding and marking; s3, determining the current optimal sparseness; s4, three different sub-populations are obtained; executing different mating pool selection strategies and offspring molecule generation strategies to obtain an offspring molecule set; s5, decoding the offspring molecular set back to a discrete chemical space, calculating an attribute value of the offspring molecular, mixing the parent molecular population with the offspring molecular population, and executing an environment selection strategy on the mixed population to obtain a next-generation molecular set; and S6, repeating the steps S2-S5 until a termination condition is met, and outputting an optimal molecular set. According to the method, the solving efficiency of the drug micromolecule multi-attribute optimization problem is ensured.

The present invention belongs to the field of security detection technology and provides a security assessment method and system for an Internet of Things malicious detector based on evolutionary computing. The present invention performs evolutionary computing on acquired malicious samples under the guidance of clustering and distance to obtain adversarial samples. During parameter evolution, if no adversarial sample corresponding to the malicious sample is found, new parameters are selected from a sequence of the number of clusters and a sequence of termination generations based on the output reason when the failure occurs and the evolutionary calculation is performed again. An adversarial attack is designed specifically for a malicious detector based on streaming text, and adversarial samples are generated using the adversarial attack method. The detector is attacked and the security of the detector is evaluated by the attack success rate, thereby solving the problem that existing assessment methods are not applicable to Internet of Things malicious detectors based on streaming text.

The invention discloses a constrained multi-task optimization method and system based on task cooperation and resource self-adaption, and belongs to the technical field of evolutionary computation and multi-task optimization. The method comprises the steps that two populations are established based on each task; a feasibility priority strategy is adopted to evolve one population, and a domain adaptation strategy based on constraint relaxation is adopted to evolve the other population; after each evolution cycle, individuals in a preset proportion are selected for evolution based on the two populations of the same task, and progenies generated by evolution are included in a progeny set; later generations generated by evolution are evaluated through environment selection, and a roulette algorithm is adopted to carry out resource adaptive allocation on an evaluation result until an evaluation counter reaches the maximum function evaluation frequency; mating is carried out based on the populations corresponding to different tasks, and generated offspring is used for randomly replacing individuals in the offspring set. According to the method, different optimization strategies are adopted to evolve and optimize different populations, the adaptability of the algorithm is enhanced, and resource waste is reduced.

This invention belongs to the field of multi-objective optimization technology, specifically involving a trend-guided dynamic multi-objective optimization evolutionary method, including: S1. Trend modeling and direction construction in intelligent scenarios; S2. Construction of perturbation and search mechanisms under trend guidance; S3. Multi-objective knee point identification and feedback reinforcement mechanism; S4. Evolutionary computation parameter control mechanism; S5. Elite sparse resampling and trend collaborative scheduling optimization. The advantages of this invention are: it constructs a closed-loop optimization system that integrates trend prediction, knee point reinforcement, perturbation generation, and adaptive feedback control, and achieves a comprehensive improvement in the convergence, directionality, diversity, and stability of the solution set in a dynamic environment, significantly enhancing the search directionality and global convergence speed, avoiding the inefficiency problem caused by relying on random perturbations, and helping to quickly locate high-quality optimal solution sets.

The application provides a multi-unmanned agent-oriented cooperative firepower attack strategy generation method, and belongs to the field of intelligent game strategy generation. The method comprises the following steps: constructing an adversarial environment, generating an initial strategy population, performing parallel calculation on the fitness of the strategy population, performing crossover and mutation on the strategy population, and iteratively evolving the strategy population. The application solves the problems of large calculation amount and long training time of the intelligent game strategy generation method, represents the strategy as an action sequence, constructs fitness based on game winning rate, uses parallel multi-opening evolutionary calculation method, and intelligently, automatically and efficiently generates a cooperative firepower attack strategy with the fitness as the optimization target.

The present invention discloses a small drug molecule optimization method based on evolutionary computation and multi-granularity proxy model approximation, comprising: 1. modeling the multi-attribute optimization problem of small drug molecules as an expensive constrained multi-objective optimization problem; 2. uniformly generating an initial molecular population in a discrete chemical space and performing an expensive optimization target evaluation and constraint evaluation on the initial molecular population; 3. encoding the evaluated small drug molecules into a continuous latent space to obtain a continuous vector representation of the molecules; 4. using the continuous vector representation of the molecules, the corresponding target values ​​and the constraint values ​​as training data to train a multi-granularity proxy model; 5. executing a proxy model-assisted evolutionary algorithm to select high-quality small drug molecules for decoding into a discrete chemical space, and performing an expensive optimization target evaluation and constraint evaluation on the initial molecular population; 6. repeating steps 3-5 to output the optimal drug molecule. The present invention aims to efficiently optimize multiple attributes of small drug molecules using a relatively low evaluation cost, thereby providing technical support for drug research and development.

The invention discloses an emergency multi-target material scheduling method and device based on a multi-task evolutionary framework. The method comprises the following steps: acquiring initial information required by emergency material scheduling; constructing a multi-task evolution framework comprising an original emergency material scheduling task and a simplified auxiliary task based on the initial information; in the multi-task evolution framework, population iteration optimization based on evolutionary computation is executed on the original emergency material scheduling task and the simplified auxiliary task, knowledge migration operation from the simplified auxiliary task to the original emergency material scheduling task is executed in the evolution process, and an optimized population of the original emergency material scheduling task is obtained; and according to a preset decision rule, determining a final emergency material scheduling scheme from the optimized population of the original emergency material scheduling task. A multi-task evolution framework is constructed, directed knowledge migration from an auxiliary task to an original task is executed, and the defects that single-task optimization is slow in convergence and direct multi-task parallel is prone to generating negative migration are overcome.

The application discloses a wireless communication signal remote interference discrimination method and device, a storage medium and a terminal, wherein the method comprises the following steps: acquiring parameter information of a to-be-detected signal, and performing normalization processing on the parameter information of the to-be-detected signal to obtain to-be-detected parameter information; inputting the to-be-detected parameter information into all interference discrimination neural networks in an interference discrimination neural network group respectively, so as to acquire interference labels output by each interference discrimination neural network; performing voting on all possible discrimination results based on the interference labels output by each interference discrimination neural network and the weights of the interference discrimination neural networks, and taking the possible discrimination result with the most votes as an interference discrimination result of the to-be-detected signal. The method of the application combines a neural network with evolutionary calculation, and realizes fast and efficient wireless communication signal remote interference discrimination with relatively high accuracy and reliability.

The invention provides a three-dimensional stratigraphic structure modeling method based on grey wolf optimization inverse distance weighted interpolation, and relates to the crossing field of geological data processing and evolutionary computation.The method comprises the steps that drilling data are analyzed, and a three-dimensional grid is constructed; constructing an IDW interpolation function, and carrying out preliminary efficiency verification on the standard parameters of the three-dimensional grid; solving a mean value by adopting a one-hole reserving normal form, and taking the mean value as a core measure of a grey wolf optimization algorithm; and obtaining refined lithologic mapping through interpolation parameter adaptive optimization driven by a grey wolf optimization algorithm, and outputting a final three-dimensional stratum model. According to the method, the three-dimensional grids are efficiently condensed from the discrete drill holes, classification loads are simplified through ordinal number mapping, KD tree injection speed-up retrieval is carried out, an optimization foundation stone is laid, and the calculation overhead is compressed.

The application discloses an abnormal sample detection and identification method and system based on evolutionary calculation and multi-modal consistency constraint, and belongs to the field of network security and artificial intelligence security. The method comprises the following steps: S1, input data collection and preprocessing; S2, multi-modal feature extraction and unified representation; S3, optimized feature selection and weight self-adaptation; S4, cross-modal fusion detection model construction; S5, abnormal sample risk identification; S6, result judgment output: generating a final detection label according to a threshold function. The application has high accuracy and strong robustness in counterfeit image and video detection and abnormal sample elimination, and can be widely applied to the fields of multimedia authentication, intelligent security and AI content tracing.

The invention relates to the technical field of electronic file anti-counterfeiting, and particularly discloses a non-inductive watermark electronic file anti-counterfeiting method based on evolutionary computation, which comprises the following steps of: S1, converting each page of document of an electronic file into a grayscale image, and generating an image sequence; s2, finding and recording optimal embedding positions, optimal rotation angles and position sequence information corresponding to a plurality of watermark images for each page of document by using an evolutionary computation method, and generating an optimal insertion position sequence of the non-inductive watermark; s3, the optimal insertion position sequence is packaged and encrypted to generate a secret key; s4, during anti-counterfeiting verification, information is inversely solved through the secret key, and whether the watermark exists or not and whether the position of the watermark is consistent or not are checked page by page by using configuration information in the secret key, so that the authenticity and integrity of the file are judged; the optimal fusion position of the watermark and the file content can be intelligently found, the imperceptible watermark is generated, verification is carried out through the encryption key, and the problems that the electronic file is easy to tamper and counterfeit, and verification is difficult are effectively solved.

A star flash encryption transmission method, apparatus and device, and a storage medium are applied to the technical field of data security, and the method applied to a first device comprises the steps of initializing a population of evolutionary computation; performing iterative optimization of a preset number of iterations on the population, and obtaining a first password set of a nonlinear iteration function value corresponding to each individual in the population of each iteration and an exponential first population fitness value corresponding to the population; according to a first password set corresponding to the first maximum population fitness value and a preset encryption function, encrypting the to-be-encrypted plaintext to obtain a ciphertext; sending the initialized population and the first maximum population fitness value to at least one second device for security verification; and if response information which is fed back by the target second equipment and indicates that the verification is passed is received, transmitting the ciphertext to the target second equipment. Data encryption is carried out by adopting evolutionary computation, so that the security and the quantum attack resistance of the encryption process are enhanced; and meanwhile, the adaptability of the encryption process is improved.

The invention discloses an image processing method and system based on branch multi-task neural network structure search, and the method comprises the steps: obtaining image data, and determining a search space which comprises all possible branch structures of a given task number in an encoder network; mask matrixes corresponding to individuals are coded through a matrix gene coding strategy to represent different network structures, and a promising branch multi-task network structure is obtained by adopting an evolutionary operator and an individual fitness evaluation strategy; a high-quality individual group is obtained through a genetic algorithm, and an optimal branch multi-task network is searched by adopting an evolutionary neural network search algorithm; training the searched neural network, and adjusting the specific weight of the task; and an optimal branch multi-task neural network structure is obtained to process the image. The evolutionary computing thought is applied to the network structure search of the branch multi-task network, the algorithm can be effectively helped to find the network structure suitable for all tasks, and it is guaranteed that multiple tasks obtain good results at the same time.

The invention discloses a self-evolutionary computing system based on dynamic logic and recursive tensor, which defines a dynamic logic system comprising an uncertainty state (U-State) and an abnormal evolutionary state (A-State), and a logic value set of the dynamic logic system is {-1, 0, + 1, U, A}. The system generates a high-dimensional solution space through iteration of a recursive tensor, when system-level contradiction is caused by exploration of a U state, an A state is triggered, then a Meta-System Receiving mechanism is started, a new rule is searched in the solution space and permanently integrated into the system, and dimension transition of system functions is achieved. According to the method, normal form transformation of a computing system from'preset instruction execution 'to'autonomous knowledge creation' is realized, and a fundamental solution is provided for solving'unknown problems' in the fields of general artificial intelligence, adaptive control systems and the like.

The invention relates to the technical field of artificial intelligence and evolutionary computing, and particularly discloses a memory annual ring learning-based evolutionary method, which comprises the following steps of: locking elite groups, extracting core declusters, coding task constraint fingerprints, backtracking efficient operators, drawing performance operator portraits, constructing memory annual ring units and storing the memory annual ring units in a memory annual ring library. When facing a new task, analyzing and generating a to-be-matched task fingerprint, matching a historical memory annual ring unit, extracting a core declustering and efficiency operator portrait, remodeling a population basis, generating a dynamic operator weight table, and executing dual-channel collaborative guide evolution; according to the method, through collaborative optimization of the solution space and the strategy space, the convergence speed and the solving performance of the algorithm in a new environment are remarkably improved, conversion from blind exploration to experience guidance is achieved, and the challenges of knowledge migration and experience reuse in dynamic change or staged tasks of a traditional evolutionary algorithm are solved.

The invention provides a sub-model weight optimization method for ensemble learning of a large language model, and relates to the technical field of data processing, and the method comprises the steps: generating an initial sub-model weight group, combining and coding the initial sub-model weight group into a binary gene character string, and taking the binary gene character string as an initial population of evolutionary computation; constructing a proxy experiment environment, generating a proxy corpus by randomly sampling the original pre-training corpus, and selecting at least three representative natural language processing tasks as proxy tasks for evaluating the adaptive value of the gene character string; based on the evaluation result of the proxy task, selecting a high-fitness gene character string by adopting a roulette method, and generating a next-generation gene character string through crossover and mutation operations; and S2 and S3 are iteratively executed until convergence, an optimal sub-model weight combination is obtained, and task-independent meta-learner fine tuning training is carried out on the original full-amount corpus to obtain a final integrated model.

This invention discloses a method, apparatus, and storage medium for automotive structure optimization. The method includes: acquiring historical data for automotive structure optimization; wherein the historical data includes structural parameters and corresponding performance simulation test results; mathematically modeling the automotive structure optimization problem based on the historical data to determine the optimization objective of the model; wherein the optimization objective is to minimize vehicle weight and maximize safety; and using a hierarchical particle swarm optimization algorithm assisted by a classification model to optimize the mathematical model and obtain structural design parameters that minimize vehicle weight and maximize safety. This invention utilizes classification model prediction to replace most of the automotive structure performance simulation during the evolution process, and uses the classification results to drive hierarchical particle swarm evolution, solving the problems of poor automotive structure optimization results and low search efficiency caused by large search space and long evaluation time in existing technologies. This invention can be widely applied in two major fields: evolutionary computation and industrial automotive structure design.

The invention relates to the technical field of resonance sound absorption regulation and control, in particular to an air layer thickness and resonance frequency correlation control method and system in a sound insulation structure. According to the method and system, a corresponding thickness range can be reversely deduced from sound pressure and structure response deviation through a Gauss-Newton inversion algorithm on the basis of the residual error minimization principle; precise positioning of a resonance interval and continuous reverse solving of thickness parameters are achieved, screening and evolutionary calculation are carried out on a multi-parameter objective function by taking a thickness correction sequence as a population through a multi-objective genetic algorithm, adaptive thickness compensation amount is generated through population iteration, and continuous integral processing is carried out on thickness difference, so that the thickness difference is obtained. The quantitative balance between the macroscopic thickness and the disturbance compensation amount is established, so that the thickness adjustment has the traceability and convergence in the time sequence, the judgment of the response slope and the full-band scanning mechanism form a feedback closed loop, and the formant position keeps stable distribution in the dynamic adjustment.

This invention relates to the field of neural network architecture search and industrial intelligent processing technology, and discloses a structure search method for multi-output dendritic neuron models for industrial classification tasks. The method includes: initializing the synaptic connection weights and dendritic threshold parameters of the multi-input multi-output dendritic neuron model to generate an initial population; uniformly dividing the initial population into several subpopulations of equal size; calculating the temporal and spatial criteria for each subpopulation after task allocation and optimization, and dynamically allocating evolutionary computational resources for each subpopulation based on the fitness change rate represented by the temporal criterion and the population distribution state represented by the spatial criterion; implementing an accelerated sharing penalty mechanism to adjust fitness values ​​according to the crowding degree among individuals to maintain solution set diversity; and updating each subpopulation. This invention solves the problem of low convergence efficiency caused by uniform resource allocation in traditional large-scale multi-objective evolutionary algorithms during neural network architecture search.

The invention discloses a self-evolution computing system and a self-evolution computing method, which realize millisecond-level online optimization and minute-level continuous evolution through a sensing-diagnosis-decision-execution-control five-layer closed-loop architecture (as shown in attached drawings of the abstract). The system has causal diagnosis, strategy self-learning and microcode level reconfiguration capabilities, the actually measured energy consumption is reduced by more than or equal to 18%, the delay is reduced by more than or equal to 25%, the decision process is completely explained, the maintenance cost of the system is remarkably reduced, and the comprehensive performance is improved.

The invention discloses a flexible manufacturing structured process multi-objective optimization method based on evolutionary computation, and belongs to the technical field of intelligent manufacturing and production scheduling. The method comprises the following steps: modeling a production problem, and taking the workload balance degree of each vehicle type as an independent objective function; performing parallel optimization on process allocation by adopting a multi-objective evolutionary algorithm to generate a Pareto optimal solution set; and visualizing the Pareto frontier through an interactive decision support module, and carrying out multi-criterion decision recommendation based on the user preference weight. According to the method, the defects that a traditional single-target optimization method cannot reflect a multi-target tradeoff relation and depends on prior weight setting are overcome, a series of high-quality scheduling schemes can be provided for multi-variety mixed-line production, and the flexibility, transparency and overall efficiency of a production system are remarkably improved.

The invention provides a shelter equipment operation control method and device based on data acquisition, and relates to the field of intelligent control. The method comprises the following steps: acquiring shelter environment data and mushroom growth data; performing time sequence alignment and fusion processing on the square cabin environment data and the mushroom growth data, and extracting square cabin microorganism behavior characteristics by using a depth time sequence model to obtain behavior characteristic vectors; transmitting the behavior feature vector to a reinforcement learning control model, outputting a control instruction, and recording a growth result after the instruction is operated; the optimized shelter parameters and operation results are uploaded to a cloud end, evolutionary computation is carried out on data from different shelters through a group learning algorithm, and an optimal parameter template is formed; and the optimal parameter template is returned to the local part of each cabin for fine adjustment, and adaptive optimization of the parameters is controlled. The method is used in a shelter equipment operation control process based on data acquisition, and solves the technical problem that an existing mushroom shelter is lack of dynamic regulation and control of operation parameters.