347 results about "Round complexity" patented technology

The invention relates to the field of nonlinear system control, and provides a preset time reinforcement learning method and system of a continuous nonlinear system, and an electronic device, and the method comprises the steps: constructing a zero-sum game framework based on a kinetic model and a performance index of the nonlinear system; determining a value function and a Hamiltonian function based on a zero-sum game framework; applying a preset neural network model to carry out approximation on the value function, and determining an approximation error; based on the Hamiltonian function and the approximation error, an approximate optimal control strategy and a worst interference strategy are determined; constructing a Lyapunov function based on the value function and the weight error of the value function; and based on a Lyapunov function, an approximate optimal control strategy and a worst interference strategy, a reinforcement learning result is verified. The method and the device are used for overcoming the defects that convergence time cannot be dynamically adjusted, parameter complexity is high and robustness is insufficient in the prior art, and the scheme of the invention can meet dual requirements of a continuous nonlinear system on dynamic convergence and anti-interference performance.

The invention discloses a virtual space multi-person interaction synchronous control method oriented to an end-cloud collaborative architecture. The invention relates to a computer graphics and three-dimensional modeling technology, and discloses a cross-format lightweight and geometric consistency maintenance method based on a three-dimensional model. Through format-independent geometric representation and a self-adaptive lightweight strategy, efficient compression and precision maintenance of three-dimensional model cross-format conversion are realized. The method specifically comprises the following steps: performing format analysis and geometric feature extraction on an input model, and establishing a unified internal representation; adaptively selecting a multi-level LOD lightweight strategy based on the complexity of the model; the accuracy of key information is ensured through geometric feature keeping and topology consistency detection; the geometric consistency is dynamically maintained by combining error monitoring and an iterative correction mechanism; and generating a target format lightweight model and carrying out quality verification. According to the method, adaptive precision control, multi-level consistency maintenance and format irrelevant processing are combined, the model size and conversion errors are effectively reduced, and cross-platform compatibility and geometric fidelity are improved. The method can be widely applied to the fields of industrial design, game development, virtual reality and the like.

The invention discloses an underwater acoustic processing CPU-GPU dynamic load balancing method based on a task flow model, and belongs to the technical field of underwater acoustic processing. The method comprises the following steps: deconstructing an underwater acoustic processing application into a task flow model represented by a directed acyclic graph; establishing a feature portrait including calculation complexity, parallelism and data throughput for each task node, and constructing a cost prediction model; the CPU / GPU utilization rate and the data transmission performance are monitored in real time; a processor is distributed to each task node by using a dynamic programming algorithm in combination with a cost prediction model and a real-time system state with the goal of minimizing the total task flow completion time; and dynamically scheduling tasks through a central scheduler according to a decision result, and periodically updating a strategy. According to the method, the defects that static task division lacks adaptability and neglects task dependence and communication overhead are overcome, dynamic and efficient utilization of CPU and GPU resources is achieved, and the efficiency and real-time performance of underwater sound processing are remarkably improved.

The invention relates to the technical field of code generation, and provides a code generation method, a power distribution network fault monitoring method, equipment, a medium and a product, and the code generation method comprises the steps: responding to a received code generation instruction input by a user, and carrying out the semantic analysis of the code generation instruction; according to a semantic analysis result, decomposing the code generation task to obtain a plurality of sub-tasks; aiming at each subtask, determining a code generation mode for executing the subtask according to the task complexity of the subtask so as to generate a code of the subtask; and generating the code of the code generation task based on the code generated for each subtask. The problem that a traditional code generation scheme is difficult to deal with a complex code generation task can be solved, the complex and multi-dimensional code generation task can be dealt with, and the code generation efficiency is improved.

The invention discloses a large language model output quality guarantee method, device and system based on chain type deterministic reasoning verification, and belongs to the field of artificial intelligence safety, deterministic calculation and reasoning verification. The method comprises the following steps: classifying each reasoning step generated by LLM into a deterministic step (bT: deductive reasoning, verified facts and mathematical proof) or a non-deterministic step (bF: inductive reasoning, unverified references and creative guess), and accumulatively constructing a reasoning chain; chain certainty verification chain (chain) = foldr (AND, true, chain) and time complexity O (n) are carried out in each k steps; if the bF is returned, a chain pollution elimination theorem is applied to prove that once the chain contains non-deterministic steps, the whole chain loses deterministic guarantee; and a deterministic boundary report is generated, and a non-deterministic initial position is marked. The device comprises an inference classifier, a reference verifier, a mathematical checker and a chain verification engine, and all the modules can be independently realized (formalized verification / database query / rule engine / mixed mode).

The invention provides a symbolic execution tree and LLM vector space heterogeneous fusion program defect detection method, relates to the technical field of program static analysis, and solves the problems of path explosion and overhigh constraint solution complexity in C / C + + program defect detection of traditional symbolic execution. The method comprises the steps that firstly, a symbolic execution tree of a target program to be detected is constructed, key feature information is extracted from the symbolic execution tree and converted into multi-dimensional feature representation, and corresponding symbolic execution feature vectors are formed; then constructing a mapping model, realizing a mapping process from the symbolic execution feature vector to an LLM vector space, and obtaining an LLM mapping result; symbolic execution analysis is achieved based on the symbolic execution tree, LLM analysis is achieved based on the LLM mapping result, and finally two kinds of analysis results are fused to obtain a program defect detection result. For optimization training of the mapping model, a comparative learning strategy is also adopted. According to the invention, accurate detection of complex program defects can be effectively realized, and the detection efficiency is improved.

The invention discloses a two-stage algorithm selection and hyper-parameter joint optimization method, which comprises the following steps of: in the first stage, processing a training set and a test set through row sampling operation and column dimension reduction operation to form a reduced data set; randomly sampling a certain number of configurations in the hyper-parameter space of each candidate algorithm, evaluating the performance of each candidate algorithm by using the reduced data set, and extracting an optimal performance score; in the second stage, a previous algorithm is screened according to the optimal performance score to form a candidate set, and a pruned hyper-parameter search space is formed so as to reduce the calculation complexity of processor hyper-parameter search; and performing hyper-parameter optimization on the pruned hyper-parameter search space by using the original data set, and outputting an optimal algorithm adaptive to the target technical task and hyper-parameter configuration thereof. Algorithm screening and hyper-parameter tuning adaptive to a specific scene are realized through a two-stage optimization strategy, and meanwhile, the method is suitable for a traditional table type dichotomy task and aims at improving the deployment efficiency and performance of a machine learning model.

The embodiment of the invention provides a model training method, an information recommendation method, equipment, a storage medium and a program product. According to the embodiment of the invention, a multi-encoder-multi-sub-decoder-total decoder hybrid model architecture is provided, sample data of different information modes correspond to different encoders-sub-decoders, and a mode of processing all training sample data by a single model is converted into a divide-and-conquer mode. The internal complexity of each encoder-sub-decoder is relatively low, the complexity of model training can be reduced, resource consumption can be saved, and different encoder-sub-decoders can be trained in parallel, so that the model training time can be shortened; and furthermore, by utilizing a dual decoding mechanism of the sub-decoder and the global decoder, parameters of the encoder can be continuously adjusted through local optimization and global optimization, the performance of the model is optimized, the accuracy of a reasoning result is improved, the convergence speed of the model is accelerated, the model training efficiency is further improved, and the model training time is saved.

The embodiment of the invention provides a model parameter compression method and device for a large language model, equipment and a storage medium. The method comprises the steps of obtaining verification data and inputting the verification data into a large language model to obtain an input activation tensor received by each network layer; obtaining the reasoning confusion degree of the large language model for reasoning the verification data, and taking the minimization of the reasoning confusion degree as an optimization target to carry out iterative cutting decision to obtain a cutting decision vector; performing outlier clipping processing on the input activation tensor of the network layer based on the clipping decision vector to obtain a target activation tensor; for each network layer, calculating an importance score of a model parameter based on the target activation tensor, and pruning the initial model parameter tensor in combination with the importance score to obtain an intermediate model parameter tensor; and performing quantization processing on the intermediate model parameter tensor of the network layer to obtain a target large language model. Therefore, the storage and calculation complexity of the large language model can be reduced while the performance of the large language model is maintained.

The invention belongs to the technical field of power grid dispatching, and particularly discloses a dual-port equivalent cluster aggregation method. According to the method, an independent aggregation operation interval is established for two ports by solving a specific robust optimization problem, an initial aggressive boundary is formed, and an initial high-dimensional polyhedral model is established. The key improvement is that the condition extreme value of the power of the other port is solved under the constraint of the operation interval of one port, so that the physical coupling relation is converted into a series of accurate boundary points, and meanwhile, the calculation complexity is also reduced. And finally, cutting the initial and aggressive high-dimensional model by using the boundaries, eliminating an infeasible space generated by neglecting coupling, and quickly generating a feasible dual-port aggregation interval. The problems of large model error and low aggregation efficiency caused by dual-port interconnection are systematically solved.

The invention provides a real-time incremental RAG method and system based on a dual-tense knowledge graph, and belongs to the technical field of artificial intelligence and knowledge graphs. The method comprises the following steps: through dual-tense decoupling modeling of event effective time and system input time, constructing a versioned knowledge graph and establishing a tense index; efficient retrieval is carried out by adopting a time sequence pruning algorithm based on indexes, and dynamic expansion of a graph pattern is realized through a declarative compiler; realizing abstract-free high-precision result sorting by utilizing a multi-modal concurrent retrieval and attention fusion mechanism; incremental updating of the knowledge graph is realized through streaming data processing and a distributed engine; and closed-loop optimization is formed according to user feedback. The system effectively solves the problems that a traditional RAG scheme is high in data updating delay, tense information is lost, retrieval efficiency is low and expansibility is poor, tense query accuracy, retrieval efficiency and system throughput are remarkably improved, and meanwhile operation and maintenance complexity is reduced.

The invention provides a CAE modeling and grid dividing method and system based on natural language driving, and the method comprises the steps: constructing a triple language model, defining an action-object-adversity triple structure, receiving a natural language instruction, carrying out the standardization processing of the instruction, recognizing and extracting the action, object and adversity components in the instruction, and carrying out the recognition and extraction of the action, object and adversity components in the instruction. The missing necessary parameters are prompted and complemented; based on a predefined analysis rule, extracting a triple conforming to the triple structure from the processed instruction; mapping the triad passing the verification into a corresponding CAE operation instruction through a mapping rule base, and generating an instruction sequence according to a logic sequence of a CAE modeling process; and automatically executing the instruction sequence, completing CAE modeling and grid division operations, and finally outputting a CAE model file. According to the method, the user operation complexity and the professional threshold are remarkably reduced, the modeling efficiency and the model quality are improved, an efficient and intelligent auxiliary tool is provided for engineering design, and the wide application value is achieved.

The invention discloses a multi-antenna anti-interference method, multi-antenna anti-interference equipment, a storage medium and a program product, and aims to solve the contradiction between poor performance of a sub-array architecture and overhigh calculation complexity of a full-array architecture. The method comprises the following steps: carrying out global synchronization processing by taking all array elements as a whole, and determining a shared synchronization point; the array is divided into a plurality of sub-arrays, each sub-array uses the synchronization point and independently calculates a set of anti-interference weights, and accordingly, sub-array output signals subjected to interference suppression are generated; and determining a group of optimal merging coefficients through an optimization algorithm with the purpose of maximizing the final signal quality, and performing weighted merging on all sub-array output signals to generate a final output signal. In addition, the method also comprises a mechanism for periodically updating an anti-interference weight and a merging coefficient. According to the method, through step-by-step processing and optimal combination, the calculation complexity is effectively reduced, meanwhile, the performance loss of a traditional sub-array scheme is avoided, and the anti-interference performance of a system in a dynamic complex environment is improved.

The invention discloses a traffic flow prediction method based on a large language model, and the method comprises the steps: data preparation, construction of a road network diagram structure, collection of traffic sensor data, construction of a traffic prediction model, formatting of the data into a mixed format in which a natural language and a structure are combined, spatial feature extraction through LLM, formatting of time series data, and reprogramming. And then fusing the spatial features with the spatial-temporal features of the reprogrammed time series data, performing lightweight fine tuning on the language model by using LoRA, then performing forward propagation, abandoning a suffix part and obtaining an output representation, performing a flattening operation on the output representation, and obtaining a prediction result through a linear projection layer. According to the method, the LLM is conveniently used for traffic flow prediction under the condition that the backbone language model is kept complete, and the complexity and parameter quantity of model training are remarkably reduced while the prediction accuracy is improved.

The invention provides a code optimization method and system based on a knowledge graph, and the method comprises the steps: carrying out the source code analysis based on a to-be-optimized code library, constructing an abstract syntax tree, and determining a to-be-optimized code snippet based on the code complexity of a corresponding code in the abstract syntax tree; determining a reconstruction mode corresponding to each to-be-optimized code snippet based on the to-be-optimized code snippet in combination with a pre-constructed code optimization knowledge graph, and constructing a context dependency graph of each to-be-optimized code snippet based on the abstract syntax tree and the code optimization knowledge graph corresponding to each to-be-optimized code snippet; performing sliding window decomposition based on the context dependency graph to obtain a plurality of mutually associated code blocks, and performing simplification and fusion on the code blocks to obtain a current context; and inputting the reconstruction mode corresponding to each to-be-optimized code snippet and the current context into the large language model for structured reconstruction to obtain a target reconstruction scheme output by the large language model. According to the method and the device, the code optimization effect is improved.

The invention discloses a dynamic beam hopping and resource allocation method for NGSO satellite security communication, and belongs to the field of sixth-generation mobile communication security communication and wireless resource allocation. According to the method, significant non-uniformity of distribution of ground flow requirements in a time domain and a space domain is considered, and a satellite-ground security communication network dynamic beam hopping and resource allocation problem model is constructed by considering ground user flow requirements and composition elements and channel characteristics of an NGSO multi-beam satellite network. Through joint optimization of satellite hopping beam scheduling, power resource allocation and auxiliary interference unmanned aerial vehicle deployment strategies, service requirements of different users and ground eavesdropping environments are dynamically adapted, and the safety throughput and queue delay fairness of the system are improved. According to the method, mixed integer linear programming modeling is adopted, and a low-complexity approximation algorithm is combined, so that the method not only has optimality guarantee, but also can be deployed and operated in an actual satellite communication system, and thus unification of secure communication and efficient resource scheduling is realized.

The invention relates to the technical field of intelligent design of optical lenses, and discloses a multi-series superposed complex periodic structure arrangement method and system.The method comprises the steps of normalized coordinate transformation, multiple sine wave superposition, an iterative function system, a Julia set algorithm, an LSystem grammar system, parameter adjustment, moire fringe fusion and the like; and a complex periodic structure with high complexity, adaptability resistance and natural morphological characteristics is generated. Advanced technologies such as logarithm uniform distribution wave number generation, a chaos game algorithm, linear mapping complex number conversion, a turn graph interpreter, multi-scale entropy complexity quantification and an Otsu automatic threshold method are adopted, and accurate parameterization control and manufacturing adaptation are achieved. The system comprises a coordinate transformation module, a wave interference calculation module, a fractal generation module, a complex power system module, an LSystem processing module, a parameter control module and a contour surface extraction module. According to the invention, a systematic solution can be provided for complex structure design of optical devices such as myopia prevention and control lenses.

The embodiment of the invention provides a multi-round dialogue complexity assessment method and device, electronic equipment and a computer readable storage medium, and relates to the technical field of artificial intelligence. According to the method, by fusing the trend of the session state, the semantic turning of the abstract and the correlation of the historical dialogues, the evaluation accuracy of the real complexity of the request in multiple rounds of dialogues is improved, and misjudgment caused by only depending on the original text length or keywords is avoided; and meanwhile, the context length transmitted to the large language model is effectively compressed by using the abstract containing intention transformation and the related dialogue, so that the Token cost is directly saved, the calculation overhead brought by redundant context transmission is reduced, and a more efficient and more intelligent model routing decision is supported. Accurate context compression and routing decision reduce the average duration of request processing, and the user experience can be improved.

The invention relates to a code generation method based on a large model dynamic routing mechanism, which comprises the following steps of: constructing a difficulty-perceived task routing module, and performing complexity judgment on an input programming problem; for simple problems, code generation is carried out by directly adopting few-sample prompt, and complex reasoning of excessive thinking is avoided; for complex problems, an intention thinking chain is introduced, reasoning is divided into two stages of specification description and thought generation, and the method is used for explicit modeling of task intention and logic structures. Then, according to a difficulty judgment result of the task, executing a generation process along the determined scheduling path: for a simple task, outputting a code by adopting a direct generation path; for a complex task, an intention thinking chain is generated firstly, and then a structured code consistent with a task intention is generated based on the thinking chain. By using the method provided by the invention, efficient generation of the programming task in the industrial intelligent system can be realized.

The application provides a uniform plane array FDA-MIMO three-dimensional target positioning method and system based on complete decoupling atomic norm minimization. The method converts the problem into a semi-positive programming solution by constructing a low-dimensional observation matrix and using complete decoupling atomic norm, and realizes joint estimation of distance, pitch angle and azimuth angle through Vandermonde decomposition. Compared with high complexity algorithms (VANM, etc.), the application significantly reduces the calculation amount by decomposing the Toeplitz matrix, and improves the single snapshot precision by using the multi-slice constraint mechanism. Compared with low complexity algorithms (3D-DANM), the application introduces an auxiliary variable matrix to restore the distance-angle coupling structure, avoids the performance loss caused by excessive relaxation, and significantly improves the estimation accuracy and multi-target resolution capability while ensuring the calculation efficiency.

The invention belongs to the field of signal processing algorithms, and provides a phase loss sparse low-rank signal recovery method based on convex programming. According to the method, through convex relaxation and anchor regression, a convex optimization frame based on an original parameter space is provided for the first time aiming at the problem of simultaneous sparse low-rank phase loss matrix signal recovery, the frame is used for recovering a simultaneous sparse low-rank matrix signal from phase loss observation data, and the frame works in an original space domain of an unknown variable. And aiming at the convex frame, a low-complexity alternating direction multiplier minimum method is developed. The sparse low-rank phase loss signal recovery method is a first sparse low-rank phase loss matrix signal recovery method based on convex optimization and in an original parameter space, and the calculation complexity is very low.

This application relates to the field of fintech and discloses a dynamic hybrid parsing and deterministic verification system and method for fuzzy bond instructions. The system constructs structured data through an instruction preprocessing module; a routing decision module calculates complexity scores based on structure, semantics, and business characteristics, and generates path selection instructions using a resource-aware dual-threshold strategy; a hybrid parsing module responds to instructions, employing a rule engine to handle low-complexity tasks and a large-scale model with enhanced retrieval to handle highly ambiguous tasks; a fallback verification module performs confidence-weighted integration and logical consistency checks, and repairs anomalies through issuer set operations and edit distance matching; a mapping generation module constructs standard query statements using an abstract syntax tree; and a closed-loop optimization module dynamically updates parameters based on user interaction feedback. This invention effectively balances computational efficiency and parsing accuracy, eliminates model illusion risks, and achieves adaptive evolution of the system.

The application belongs to the technical field of electric data processing, and particularly relates to a nonlinear unit construction method based on dimension expansion. The shell unit nodes are classified and information is stored; the normal of the shell unit is calculated; new nodes are expanded on the positive and negative sides along the normal according to the thickness, and corresponding three-dimensional solid units are generated; the junctions at the unit connections are specially processed to maintain the connectivity; and the original constraint conditions are converted and applied to the expanded new nodes. The method significantly improves the calculation accuracy and boundary adaptability, greatly reduces the development complexity of the nonlinear unit, and facilitates the direct coupling analysis of beams, shells, solids and other units in a unified three-dimensional system.

The application discloses a heterogeneous platform approximate computing task optimization mapping method based on DVFS and DPM. Firstly, real-time tasks with correlation are modeled as an approximate computing task model, so that a task directed acyclic graph (DAG), a task correlation matrix and a six-tuple representing task characteristics can be obtained; then, a mechanism of DVFS and DPM combination is introduced based on a heterogeneous multi-core platform; a problem description of task mapping based on QoS and energy joint optimization is constructed; a variable substitution method and a Big-M reconstruction method are used to process nonlinear terms in the problem, the task mapping problem is linearized, and an optimal solution is obtained through a Gurobi solver; a task layering method and a greedy algorithm are used to design a low-complexity heuristic algorithm, and the scalability of the mapping method is improved. The method of the application adopts the mechanism of DVFS+DPM joint optimization under the premise of meeting the system real-time, energy efficiency and reliability constraints, and improves the QoS of the system.

The invention discloses a dynamic motion primitive coding method and system based on Lie group Lie algebra, and the method comprises the steps: uniformly representing the tail end pose and contact force spinor of a robot in a Lie group form, and constructing a unified mathematical framework; the method comprises the following steps: acquiring a pose and a force spinor sequence through a teaching phase, and converting the force spinor sequence into a virtual pose sequence through scaling mapping; establishing a dynamic motion primitive (DMPs) model based on the Lie group Lie algebra theory, and learning a model weight matrix and a primary function parameter by adopting local weighted regression; and according to the initial and target poses of the new task, generating a smooth trajectory through phase synchronization control and Euler integration, and outputting the smooth trajectory to a robot controller. According to the method, the problems of posture singularity and motion distortion caused by splitting of a posture channel in a traditional DMPs method are solved, the learning fidelity and reproduction stability of the six-degree-of-freedom posture track of the robot are improved, the smoothness and coordination of the whole motion process are guaranteed, and the debugging complexity of a force control system is reduced.

The invention provides an application program execution method and system and a storage medium. The method comprises the steps of splitting an application program into a plurality of step functions by calling a plurality of preset functions, and obtaining corresponding message data based on a preset message queue; constructing a flow table and an application execution table based on the plurality of step functions and the message data; traversing the application execution table through the message data to obtain a corresponding flow table; and acquiring a corresponding step function based on the flow table so as to execute the step function in a time-sharing manner. According to the method provided by the invention, the technical problems that when the multi-task management is carried out in the real-time operating system, the execution time of the task is relatively long or the calculation complexity is relatively high, so that the thread is blocked, the blocked task occupies system resources, other tasks are delayed or stopped, and even the system function is abnormal are effectively solved; the complexity of the application program is effectively simplified, and the execution efficiency of the application program is improved.

The invention discloses a topological optimization method and device considering volume constraints, boundary conditions and a design domain based on deep learning, and belongs to the technical field of topological optimization design. According to the method, training data sets under various constraint conditions and the design domain are constructed, a full convolutional neural network is adopted for offline training, and the training data sets are optimized; and obtaining a full convolutional neural network model capable of representing a topological optimization design rule. In practical application, only corresponding boundary conditions and design domain information need to be input, a topological optimization result can be rapidly output through the trained neural network, and therefore a traditional iterative calculation process is replaced. According to the method, the calculation complexity and the time consumption are remarkably reduced while the topological optimization precision is kept, high efficiency and intelligence of topological optimization are realized, and the method has a good engineering application prospect.

The invention provides an APT attack detection method, device and system based on a lightweight deep learning framework. The method comprises the following steps that: each node constructs a local sample data set based on local network flow data to train a local classification model, and uploads trained model parameters to a federated server; the federal server aggregates the model parameters of all the nodes and optimizes global model parameters, so as to issue the optimized global model parameters to each node; performing parameter updating on the local classification model by each node based on the received global model parameter, and repeatedly executing the model training step on the updated local classification model until the model converges; and each node classifies the to-be-detected traffic data based on the converged global model parameters to obtain an APT attack detection result. According to the technical scheme provided by the invention, the complexity and communication load of the model are effectively reduced while the detection precision is ensured, and the practicability and stability of the model in a distributed and resource-constrained environment are improved.