250 results about "Adaptation method" patented technology

The invention belongs to the technical field of intelligent control, particularly relates to a deep learning-driven intelligent home scene dynamic adaptation method, and aims to solve the problem that an existing intelligent home system is difficult to realize high-precision personalized scene adaptation in a multi-user and multi-device environment due to dependence on a static rule. The method comprises the steps of collecting multi-source heterogeneous user behavior data and performing semantic enhancement preprocessing, constructing a hierarchical time sequence behavior coding model to extract local time sequence dependence and cross-equipment long-range association features, clustering to generate a dynamic scene prototype and mapping the dynamic scene prototype into an executable condition-action rule, after the rules are deployed, a closed-loop optimization mechanism is constructed through explicit and implicit user feedback, and online incremental updating and self-adaptive evolution of the behavior model and the scene rules are achieved. According to the technical scheme, the user complex behavior mode can be deeply understood, the scene adaptation precision is continuously optimized, the individuation level, logic consistency and system robustness of intelligent services are improved, and meanwhile privacy safety and real-time response are guaranteed through edge calculation.

The invention discloses an intelligent routing and unified adaptation method for a large language model, and the method comprises the steps: defining all access details of the model through a declarative configuration file, and achieving the zero-code access of the large language model without writing any code for a newly-added model; a completely consistent calling interface is provided for an upstream application, the isomerism of all downstream large language models is shielded, and a unified request and response abstraction layer is constructed; through a strategy engine and a JSON path technology, complex streaming response including content thinking is precisely processed, and intelligent analysis and content extraction are carried out; dynamic configuration and intelligent strategy hot update are supported; intelligent routing of the model is realized, and an optimal large language model instance is dynamically selected; and meanwhile, enterprise-level governance capability is provided, governance functions such as fusing, degradation, current limiting and monitoring are integrated, and stability guarantee is provided for model calling. Therefore, the maintainability, the expandability and the user experience consistency of the system are comprehensively improved.

The invention relates to the field of artificial intelligence, in particular to a heterogeneous GPU-oriented large model reasoning platform adaptation method and system. The method comprises the following steps: continuously and dynamically collecting and maintaining all available GPU resources in a server cluster and attribute and state information of a deployed reasoning acceleration framework instance through proxy service deployed in each server node; providing a single entry point to receive a reasoning service request, and analyzing the request to determine a large model needing reasoning; according to the analyzed large model needing reasoning and the currently aggregated resource information, selecting a reasoning acceleration framework instance which is optimal in load state and compatible for the request through a decision algorithm, and routing the task to the reasoning acceleration framework instance; heterogeneous response formats returned by different inference acceleration framework instances are converted and packaged into a unified standardized format, and then the unified standardized format is returned to a requester. Scattered and diversified computing resources are effectively utilized, and dependence on a single hardware manufacturer is reduced.

The invention relates to an industrial protocol adaptation method, system and device based on dynamic feature recognition and a medium. The method comprises the following steps: acquiring original communication data of industrial equipment; performing feature extraction on the original communication data to obtain a depth feature vector; performing feature matching on the depth feature vector in a known protocol feature library to obtain a protocol matching result; if the protocol matching result is the unknown protocol parameter, marking the depth feature vector as unknown to obtain an unknown feature vector; clustering all the unknown feature vectors based on the protocol matching result to obtain candidate protocol groups; and for each group of candidate protocol groups, generating a new protocol analysis rule set. The method can support parallel operation of multiple communication protocols, is suitable for different manufacturer devices, automatically classifies unknown protocols through a machine learning algorithm, and enhances self-learning and self-adaptive capabilities.

The invention provides an AI flow chart driven test step dynamic adaptation method, device and equipment, and relates to the technical field of software testing. The AI flow chart is subjected to semantic analysis to obtain a plurality of executable operation tuples, a dynamic equipment capability library is updated in real time through technical parameters and real-time states of all the matching and testing instruments, then dynamic matching and mapping are carried out, a first mapping relation between the executable operation tuples and the matching and testing instruments is established, and the dynamic equipment capability library is updated in real time through the technical parameters and the real-time states of all the matching and testing instruments. According to the second mapping relation between the signal features and the port of the matching and testing instrument, the AI flow chart is executed based on the executable mapping scheme, automatic matching and mapping of all the testing steps in the AI flow chart are achieved, manual participation is not needed, the problem that the testing efficiency of AI flow chart driving is low is solved, and the equipment testing efficiency of AI flow chart driving is improved.

The invention provides a gear tooth surface precision detection method based on a deep learning network, a medium and equipment, and mainly solves the problem of insufficient model generalization ability caused by scarcity of real labeled data and domain difference in an existing method. The method comprises the following steps: collecting a plane workpiece image marked with a roughness level as a source domain data set, and pre-training an intelligent detection model; based on a gear shaping machining meshing motion relation, establishing a tooth surface morphology geometric simulation model, and generating virtual gear tooth surface images of different roughness grades; extracting a real gear tooth surface area image as a target domain sample; a domain adaptation method based on clustering optimal transmission is adopted, the feature distribution difference between a source domain and a target domain is calculated and minimized, model parameters are optimized, and self-adaptive migration of a feature space from a plane image to a gear tooth surface is achieved. According to the method, virtual sample generation and a domain adaptation technology are combined, so that the data acquisition cost is effectively reduced, and the detection precision and the model generalization ability under the small sample condition are remarkably improved.

The invention provides a low-delay video stream code rate adaptation method and system based on off-line element reinforcement learning. The method comprises the following steps: constructing an off-line expert track data set; constructing a code rate adaptation strategy network, and outputting a downloading code rate selection result and a network throughput correction value by taking a video stream transmission state as input; and adopting a meta-implicit Q learning algorithm to perform double-layer alternate optimization of an inner layer and an outer layer on the policy network, the inner layer updating policy parameters of the policy network based on the offline expert trajectory data set, and the outer layer further adjusting the policy network parameters through meta reinforcement learning to improve the adaptability of the policy network to unseen tasks. According to the method, multiple expert algorithm experiences are integrated through offline reinforcement learning, high generalization and adaptive ability can be realized in multiple target time delays and complex network environments without online trial and error exploration, and finally, high-quality and smooth playing of video streams in a low-time-delay live broadcast scene is supported.

The invention provides a dynamic adaptation method and system for a multi-mode power supply compatible charging pile system, and the method comprises the steps: building a multi-protocol physical layer compatible connection between a charging pile and a power supply module through a multi-communication interface layer, and receiving a data frame inputted by the charging pile; analyzing a baud rate, a check bit and an instruction set of the data frame, and identifying a charging pile protocol type and a power module protocol type; querying a matching mapping relationship between the charging pile protocol type and the power supply module protocol type in the dynamic adaptation matrix; if the matching mapping relation exists, data field conversion is executed, and a target protocol frame is generated; and if the matching mapping relation does not exist, starting a machine learning process, creating a temporary mapping rule, and effectively solving the problems of poor compatibility and high manual configuration cost of the charging pile and the power supply module in a traditional protocol mode through the modes of protocol analysis, intelligent matching and dynamic learning.

According to the intelligent terminal equipment intelligent adaptation method and device, intelligent terminal automatic discovery is carried out based on scene-based adaptive ARP scanning and multi-fingerprint UPnP filtering, access equipment is obtained, automatic and precise discovery of intelligent terminal equipment is achieved, and the intelligent terminal equipment can be automatically and accurately found by standardizing the scanning frequency and the filtering rule. The equipment discovery efficiency and accuracy are improved, the access equipment is subjected to prefabricated API library matching and AI reverse analysis to obtain interface information of the access equipment, automatic analysis of a standard interface and a private interface is realized based on a multi-dimensional analysis technology of an AI large language model, and the equipment discovery efficiency and accuracy are improved by utilizing the AI large language model and combining a built-in multi-language programming template. According to the method, the adaptive code matched with the interface information is automatically compiled, the adaptive code is automatically generated based on the AI technology, the compatibility of multi-protocol and multi-interface type equipment is supported, and finally the core problems of low efficiency, poor compatibility, difficult maintenance and high threshold in the existing intelligent terminal adaptation process are solved.

The invention relates to an accelerator adaptation method for AI algorithm reasoning, and the method comprises the steps: carrying out the structural analysis of a to-be-deployed AI reasoning model, recognizing a calculation path having an influence on a reasoning result based on a target reasoning output result, and cutting or recombining a model structure, so as to generate a group of subtask sets for representing main reasoning behaviors; according to a basic calculation unit structure of a target accelerator, the subtask set is subjected to modular division, the calculation granularity of each module is made to be matched with an execution unit supported by the accelerator, and the operation sequence between the modules is adjusted; aiming at the execution path of the modular task, analyzing a resource mapping relation on a target accelerator, and reorganizing a calculation path execution sequence located in a resource bottleneck region; based on the execution state of the current accelerator, including thread activeness, cache use condition and resource occupation condition, an adaptive instruction sequence is dynamically generated and optimized, and multi-level adaptive optimization is scheduled from a model structure level to an instruction level.

The invention relates to an interface adaptation method and system for reconstruction and integration of a complex information system module. The method comprises the following steps: acquiring a protocol mapping rule and constructing a rule base; monitoring and receiving a calling request which is sent by a source module through a transmission layer protocol and carries source module side interface protocol format data; analyzing the request, stripping a source module side transport layer protocol header to extract application layer data, and packaging the application layer data into an internal general data model; performing structure mapping, data conversion and dynamic logic processing on the model through a rule execution engine according to a request mapping rule of a rule base, reconstructing the model into target request data according to a target module interface format, and sending the target request data to a target module through a target side transport layer protocol; receiving target response data, stripping a transport layer protocol header of the target response data, analyzing the target response data to the internal general data model, and performing reverse conversion according to a response mapping rule to enable the data to conform to a source side interface format; and packaging the back-to-source module side transport layer protocol back to the source module. By adopting the method, the adaptation flexibility and the maintenance efficiency can be improved.

The invention provides a rapid database adaptation method and device, a storage medium and electronic equipment. The method comprises the steps that an execution request which is initiated by an application program and contains a source SQL is obtained; enhancing the database execution object to obtain an enhanced object; intercepting the source SQL through the enhanced object; performing structured analysis on the source SQL to generate a corresponding intermediate representation structure; loading a corresponding dialect conversion rule set according to the type of the target database; generating a target intermediate representation structure according to the dialect conversion rule set; according to the target intermediate representation structure, rendering and generating a target SQL (Structured Query Language) compatible with a target database grammar; calling a native drive of the target database to execute the target SQL to obtain a native execution result; performing compatibility packaging processing on the native execution result to obtain an adaptive execution result; and returning an adaptive execution result to the application program. According to the method, large-scale modification of business codes and SQL statements is not needed, and the migration cost and risk are effectively reduced.

The invention provides a desktop browser plug-in compatible adaptation method and system based on a credential terminal Web application, and relates to the technical field of credential terminal Web application, firstly, behavior data of a plug-in during operation of a credential terminal is captured, including calling a system interface, accessing local resources and outputting a data rendering record, and time sequence correlation analysis is performed on the data; the method comprises the following steps: establishing an interaction behavior fingerprint, identifying an abnormal interaction node, calling a mismatching pattern library to determine a mismatching reason and generate an adaptive adjustment instruction for the abnormal node, injecting the instruction into a plug-in running process, starting a behavior monitoring process, recording the adjusted interaction behavior fingerprint, and judging a plug-in function completion condition based on adjusted data. The effective adjusting parameters are extracted to construct the adaptation operation process, the final compatible adaptation scheme is generated, and the compatible adaptation efficiency and accuracy of the plug-in and the credential terminal can be improved.

The invention relates to the technical field of non-intrusive load monitoring, in particular to a domain adaptation method and system for non-intrusive load monitoring. Comprising the steps of obtaining source domain synthetic data and target domain real data; constructing a domain separation network, wherein the domain separation network comprises a shared encoder, a private encoder, a decoder and a regression device; public features and private features are extracted from the source domain data for reconstruction and energy consumption prediction; fusing features of the target domain data through an external attention mechanism to obtain enhanced features and reconstructing the enhanced features; a total loss function is constructed, characteristic orthogonal decoupling of the difference loss is realized through a Floribenius norm, and distribution alignment of the similarity loss is realized through domain adversarial training; a gradient descent method is adopted to optimize parameters, and when the iteration step number reaches a threshold value, a gradient inversion layer is introduced to carry out domain confrontation training. According to the method, the distribution difference between the synthetic domain and the real domain is effectively relieved, negative migration is avoided, and the cross-domain generalization ability is improved.

The application provides a heterogeneous computing power automatic adaptation method and system for an intelligent chip, relates to the technical field of heterogeneous computing adaptation of the intelligent chip, and acquires computing graph structure and operator feature data and hardware state data of a to-be-deployed model; an execution knowledge base is constructed, a scheme list is generated in combination with real-time data and the execution knowledge base; execution code is distributed according to the decision weight of the scheme list; the current execution state is evaluated at a predefined execution decision point to determine whether to switch the adaptive path; the actual execution effect and environment data generated after each decision are summarized, execution track records are formed, and the execution knowledge base is updated to periodically optimize the basis for scheme generation. The application can realize automatic perception, decision, optimization and online learning of a computing task in a heterogeneous computing power environment for an intelligent chip, significantly reduces the adaptation cost and improves the resource utilization efficiency.

The application discloses a water conservancy knowledge differentiation retrieval and query decomposition method based on a large language model, and the method comprises the following steps: constructing a standardized retrieval channel for four types of heterogeneous data sources, namely a graph database, a relational database, a document library and a time series database; secondly, establishing a sub-query type classification system for the water conservancy field and forming a template library; thirdly, collecting training data from historical queries and expert samples, and adopting a multi-round labeling mechanism to label query decomposition results, channel mapping relationships and dependency relationships; finally, adopting a low-rank adaptation method to fine-tune the pre-trained large language model, so that the model can intelligently decompose a user natural language query into a sub-query sequence, automatically route to the corresponding retrieval channel and plan an execution sequence. The application realizes intelligent processing of a composite query in the water conservancy field, significantly improves the efficiency and reliability of cross-data source knowledge retrieval, and is suitable for decision support scenes such as flood control scheduling and reservoir group joint scheduling.

The present application relates to the field of data processing, and discloses a multi-parameter fusion load balancing scheduling adaptation system and method, wherein the multi-parameter fusion load balancing scheduling adaptation method comprises: predicting the demand and the actual available capacity in the future period in a rolling time domain, and calculating the disturbance intensity; constructing a resource department network, calculating the influence intensity and the comprehensive disturbance influence; calculating the over-allocation degree, and constructing a conflict over-allocation graph; generating an initial feasible solution; constructing a multi-objective optimization model, and solving to obtain a resource allocation scheme; the present application can accurately identify and quantify the propagation effect of the disturbance in the resource department network by establishing a three-layer progressive conduction mechanism, realizes accurate prediction and evaluation of the disturbance influence; the problem decomposition is realized through the Lagrange relaxation framework, can handle large-scale scheduling problems containing complex rules, realizes multi-parameter fusion load balancing scheduling adaptation in a cross-department collaborative scenario, and provides a complete technical solution for enterprise-level resource management.

The application provides an adaptation method and device and a computer readable storage medium, relates to the technical field of application software, and can provide functional support for automatic code replacement execution and scheme between source adaptation content and target adaptation content. The method comprises the following steps: when an adaptation application request is submitted, target parameters are configured, including an adaptation environment, basic data, a code library address and a function list; in the case of completed configuration, the adaptation application request is submitted to an automatic adaptation service through a form; the code library address information is transmitted to a machine learning service through the automatic adaptation service, and machine learning training is performed through the machine learning service; the function set in the code file information and adaptation requirements are used for rule matching on an adaptation rule library; in the case of successful matching, a code replacement operation is performed on user code; and based on the file name and file position information of the replaced user code, a code submission request is triggered. The application is applied to automatic code replacement execution.

The invention discloses a protocol dynamic adaptation method for a stock power Internet of Things terminal, and relates to the technical field of power Internet of Things and information communication computers. Comprising the steps of collecting communication message data and generating a standardized message data set in combination with a protocol mode analysis method based on an interaction process of side equipment and a stock power Internet of Things terminal; performing preprocessing based on the standardized message data set, and performing feature extraction to form a feature set adapted to model training; constructing a protocol recognition model according to the feature set, and realizing online incremental training and self-adaptive modeling of a newly added unknown protocol through a dynamic updating mechanism; and deploying the protocol identification model to an edge side, performing reasoning identification on stock power Internet of Things terminal communication message data acquired in real time, and if an identification result is not in a local registry, triggering a federal learning update process to realize protocol dynamic adaptation and continuous analysis of the stock power Internet of Things terminal.

The application discloses an AI-driven game 3D asset modular generation and multi-terminal intelligent adaptation method and system. The method comprises the following steps: firstly, a multi-task deep learning model integrating semantic segmentation and graph convolution network is used to perform structural feature recognition and disassembly on an input 2D game design drawing, and a structured descriptor containing component semantics, key points and topological relations is generated. Secondly, a generative AI model is used to generate a parameterized 3D geometric body based on the descriptor, and the parameterized 3D geometric body is encapsulated into an independent asset module with a standardized connection interface to construct a searchable modular asset library. Finally, according to a target terminal performance knowledge graph, a reinforcement learning-based adaptation agent is used to automatically optimize the rendering parameters of the modular assets, and a closed-loop verification of performance and visual consistency is performed in a simulation environment, and an adapted asset package is output.

The application relates to a STEAM teaching robot environment adaptation method based on deep reinforcement learning, which comprises the following steps: S1: collecting classroom multi-modal environment data and preprocessing, and obtaining unified multi-modal data after preprocessing; S2: constructing a classroom semantic map and an environment change factor; S3: multi-dimensionally modeling a student state, and obtaining a student state matrix; S4: constructing a unified state representation processable by a reinforcement learning model; S5: according to the unified state representation processable by the reinforcement learning model, generating a robot teaching adaptation strategy through a hierarchical deep reinforcement learning model, and obtaining a candidate action; S6: performing safety and teaching suitability checking on the robot action, obtaining a robot finally allowed action, and sending the robot finally allowed action to a robot motion control module, a voice interaction module and a teaching content display module. The application effectively improves the classroom perception ability and autonomous adaptation ability of the STEAM teaching robot.

The application discloses a flash memory partition identification multi-hardware automatic adaptation method and system, relates to the technical field of embedded system starting, and steps are as follows: presetting hardware type information identification and data check information at the device end, and writing the information into a predefined partition according to a preset format; storing a device tree binary file into a resource image according to a uniform rule; after power-on of the device, a bootloader mounts the predefined partition, automatically matches and extracts the hardware type information identification and check information and checks; if the check is passed, corresponding device trees are parsed, matched and loaded, if the check is not passed or there is no matching file, a preset default device tree is loaded; and the device trees are transmitted to an operating system kernel to complete initialization. The application solves the problems of device tree dynamic selection difficulty, many firmwares in traditional schemes and high operation and maintenance costs in the scene without physical hardware identification signals. The application achieves the goal of no additional hardware and one set of firmware compatible with multiple hardware versions, improves reliability and expandability, and reduces production and operation and maintenance costs.

The invention discloses a multi-modal environment dynamic adaptation method based on balance dysfunction of Alzheimer's disease, which comprises the following steps: acquiring three-dimensional attitude image data of a patient through an RGB-D camera, synchronously acquiring plantar pressure distribution of a pressure sensor array, and forming a composite original data set; processing data in real time by adopting an edge computing node, extracting gait key point coordinates and computing a gait stability index; performing multi-modal fusion on the indexes and the environmental parameters to construct a digital twinborn model, dynamically adjusting modal weights through federated learning, and outputting obstacle early warning; and the tactile feedback waistband activates the motor according to early warning grading and triggers an environment parameter adjustment strategy. According to the method, full-link dynamic adaptation from data collection to environment feedback is achieved, the patient balance ability is effectively improved, privacy data are protected, and the method has remarkable technical innovation and clinical application prospects.

The invention discloses a sign language translation test adaptive method based on a sign language person perception hybrid expert network, and aims to solve the problems of poor field offset adaptation instantaneity, insufficient individuation and weak stability in a real scene in the existing sign language translation technology. According to the method, a pluggable sign language perceptual hybrid expert (SAME) module is embedded in a pre-trained sign language translation model encoder, and online self-adaption is realized in combination with unsupervised adaptation loss; the SAME module adopts a low-rank expert design and sign language perception gating mechanism, and expert collapse is avoided through distribution loss initialization. According to the method, the target domain does not need to label data, self-adaptive learning can be carried out on the new domain during testing, the performance of the pre-training model is improved, and the model is helped to adapt to dynamic environment changes of cross-sign language persons.

The application relates to a visual language model test adaptation method based on logarithmic calibration and consistency cache, and relates to the technical fields of computer vision, pattern recognition, machine learning and artificial intelligence. The method aims at the problems of class prediction bias, insufficient cache sample coverage and low utilization rate of boundary samples of a visual language model in a target domain test stage, and constructs an online adaptation framework composed of an image encoder, a text encoder, a dynamic logarithmic calibration module, a consistency guide exploration cache module and a cross-modal joint optimization module. The method improves the identification opportunities of difficult classes and low-frequency classes through a dynamic logarithmic calibration mechanism, and improves the overall class balance and classification stability. Through the consistency guide exploration cache mechanism, the coverage range of the cache to the real distribution of the target domain is expanded, and the perception ability of the model to the decision boundary region is enhanced, so that the adaptation effect in a complex distribution offset scene is improved.

Methods, systems, and computer-readable storage media for machine learning database memory use prediction and adaptation. An example method includes determining a sampling interval for an application for sampling memory use by a database for the application. A plurality of historical memory use samples of amounts of memory used by the database are determined for the application based on the sampling interval. The plurality of historical memory use samples are provided for training of a machine learning model to predict memory use for the application by the database for a future time period. A set of current memory use samples are provided to the machine learning model and a memory use prediction for the application for an upcoming time period is received from the machine learning model. A determination is made as to whether to extend memory of the database for the application based on the memory use prediction.

The invention relates to a spatial position instruction fine tuning method based on a multi-modal large language model, and the method comprises the following steps: S1, converting a spatial position reasoning data set into a visual instruction format through employing a dialogue template, and obtaining a visual spatial position reasoning data set; s2, acquiring a large language model InternVL as a multi-modal large language model, performing pre-training on the general data set to obtain a pre-training model, reasoning the data set based on the visual spatial position, adjusting parameters of the pre-training model by adopting a low-rank adaptation method to obtain a trained large language model, and outputting a description corresponding to a spatial task by the large language model; and S3, introducing a text-based large language model, and optimizing the description corresponding to the space task based on the large language model. Compared with the prior art, the method has the advantages that the ability of the multi-modal large language model in understanding and generating context rich description is fully utilized, and the ability of the model in generating accurate and detailed description is enhanced.

The embodiment of the invention provides an API (Application Program Interface) protocol adaptation method and system, electronic equipment and a storage medium, and relates to the technical field of computers. According to the method, semantic understanding can be carried out on a description file of a first API protocol on a client side and a description file of a second API protocol on an API server side through a generative agent; the first API protocol is used for identifying the protocol content difference of the first API protocol compared with the second API protocol, and automatically generating a corresponding adaptation layer code according to the identified protocol content difference; and an adaptation layer code generated by the generative agent is loaded to the gateway, so that the gateway converts the API request sent by the client into a format adaptive to the API server by means of the adaptation layer code. Through application of the embodiment of the invention, the problem of compatibility between the server and the API server can be solved, meanwhile, the manpower cost and the time cost consumed during API protocol conversion are saved, and the adaptation period between different API protocols is shortened.

The invention relates to a flight control program intelligent adaptation method and system oriented to a heterogeneous processing architecture, and solves the technical problems of low efficiency, strategy solidification, difficulty in knowledge reuse and the like when a flight control program is manually transplanted to a domestic or novel chip in the prior art. And automatically generating a customization strategy containing environment configuration, a peripheral list, a test path and branch planning. Accordingly, a debugging environment is automatically deployed, a drive code is generated, and parallel development branches are initialized. An event-driven and dependent sensing mechanism is adopted, verification tasks are dynamically scheduled, code integration is coordinated, hardware abstraction layer configuration and interface codes are intelligently adjusted, a flight control program image is finally and automatically constructed, and efficient and reliable heterogeneous chip flight control program intelligent adaptation is achieved. The method has the following technical effects that automatic intelligent adaptation of the flight control program to the heterogeneous chip is realized, the transplantation efficiency and the code quality are greatly improved, and reusable adaptive knowledge assets are formed.

This invention discloses a method and system for intelligent operation adaptation in cleanrooms based on environmental constraint linkage, relating to the field of cleanroom operation and adjustment technology. It includes: S1, real-time acquisition of multi-constraint operation data and data preprocessing; S2, analysis of the changing characteristics of the multi-constraint operation data, assessment of the urgency of environmental constraints, generation of multi-constraint urgency comparison results, determination of the current dominant constraint type, and output of priority control strategies; S3, using the changing characteristics of the multi-constraint operation data to determine the conflict between energy consumption costs and pressure difference benefits, adjusting the supply and exhaust valve ratio, fan frequency increase, and filter unit operating level; S4, monitoring the operating frequency and changing characteristics of the supply fan, assessing the stability of the linkage control, and implementing optimized adjustment actions. This solves the problems of fixed priorities being difficult to adaptively adjust under dynamic multi-constraint conditions in cleanrooms, easily leading to constraint conflicts and mismatches, resulting in energy consumption fluctuations and operational instability.