90results about How to "Efficient sharing" patented technology

The invention provides a data fusion sharing method for multi-source heterogeneous power big data, which comprises the following steps: when receiving multi-source heterogeneous power data, carrying out dynamic protocol adaptation, and converting multi-source heterogeneous power data formats of different protocols into a current system internal standard format; performing data quality evaluation on the data converted into the standard format in the current system, and when a quality evaluation result does not meet a quality requirement, performing data restoration; analyzing the data passing the data quality evaluation or restoration, determining the correlation among different data sources, and fusing the data according to the correlation among the different data sources; and performing distributed storage on the fused data for distributed indexing and access. By implementing the method and the system, efficient fusion and sharing of multi-source heterogeneous data can be realized, the intelligent level of data management and analysis of a power system is improved, and the requirements of an intelligent power grid on data instantaneity, accuracy and safety are met.

The invention relates to the technical field of data processing, in particular to a real estate surveying and mapping data processing method and system based on intelligent data. The method comprises the following steps: acquiring multi-source real estate surveying and mapping data; performing space-time synchronization on the multi-source real estate surveying and mapping data to generate multi-source real estate surveying and mapping space alignment data; real estate space feature extraction is carried out on the multi-source real estate surveying and mapping space alignment data to obtain real estate space feature data; performing cross-domain feature fusion on the multi-source real estate surveying and mapping data based on the real estate spatial feature data to generate real estate surveying and mapping fusion spatial feature data; public surveying and mapping point confirmation is carried out on the multi-source real estate surveying and mapping data through the real estate surveying and mapping fusion spatial feature data, and real estate surveying and mapping public surveying and mapping point data are obtained. Through multi-source data space-time synchronization, feature fusion, distortion correction, high-dimensional space reconstruction, semantic tag association and chain storage, the efficiency and accuracy of real estate surveying and mapping data processing are improved.

The invention relates to the technical field of AUVs (Autonomous Underwater Vehicles), in particular to a real-time communication and positioning data fusion processing method for AUV (Autonomous Underwater Vehicle) collaborative operation, which comprises the following steps of: deploying a multi-modal sensor, acquiring communication data, kinematics data, acoustic positioning data and visual data required by an AUV, and preprocessing the data; the method comprises the following steps: fusing multi-modal data by using an extended Kalman filtering algorithm, calculating three-dimensional position and attitude information of the AUV in real time, carrying out dynamic path planning by using an ant colony optimization algorithm based on fused positioning data, optimizing a collaborative operation path between the AUVs, and sharing the positioning data and a path planning result in real time through an underwater communication network. And dynamically adjusting a task allocation strategy by using an improved task allocation algorithm, optimizing the cooperation efficiency, and finally, storing all data, path planning records and task allocation schemes in a database. According to the invention, the accuracy and stability of AUV collaborative operation are improved.

The invention relates to the technical field of disaster prediction and early warning, in particular to a rural water supply project drought disaster prediction and early warning method and system. The method comprises the following steps: acquiring a hydrological data set; performing data preprocessing on the hydrological data set to generate a hydrological core processing data set; performing data feature extraction on the hydrological core processing data set to generate a hydrological feature data set; quantifying the hydrological characteristic data set to generate a multi-water-source drought early warning threshold calculation algorithm; constructing a multi-dimensional interaction model by using a multi-water-source drought early warning threshold calculation algorithm in combination with the hydrological characteristic data set, and generating a multi-source hydrological drought early warning and prediction model; therefore, by optimizing hydrological data processing and early warning threshold calculation, the problem of inaccurate prediction caused by inconsistent multi-source data in a traditional early warning system is solved, and the accuracy and reliability of early warning of drought disasters are greatly improved.

The invention provides a data access control method and device based on a block chain and improved CP-ABE, and belongs to the technical field of block chains. The method comprises the following steps: an authority generates key parameters including a system public key and a master key, and uploads the master key to a block chain; the user sends an attribute set and an identity label of the user to an authority, the identity label of the user is hidden, an attribute private key, a search private key, a conversion key and a recovery key are generated for the user, the conversion key is sent to the cloud server, and the attribute private key, the search private key and the recovery key are sent to the user, and the user comprises a data owner and a data visitor; the data owner carries out offline encryption, online encryption and index encryption on the file in sequence, and index information is linked; the data accessor performs data access; and the data accessor decrypts the data. According to the method, the problems in data storage and access control are effectively solved, the integrity of user data is protected, and the calculation burden of a data access user is reduced.

The invention relates to a multi-source remote sensing data quantitative inversion regional crop water consumption demand data fusion method. The method comprises the steps of data acquisition, data assimilation, data fusion and formation of a multi-source heterogeneous database. A distributed multi-source heterogeneous database is built by using advanced technologies such as comprehensive grid computing and cloud computing, a high-performance storage retrieval multi-source heterogeneous database and a data product oriented to a complex irrigation data system are built, agricultural information is collected in real time, a database is formed, and a database management system is established. And finally, efficient data management and sharing are realized. Related agricultural irrigation processes are systematically monitored, including monitoring of water consumption of crops, monitoring and forecasting of agricultural irrigation and the like. An agricultural production management system and an agricultural expert system are developed and are mainly used for diagnosing water shortage, fertilizer shortage and pest and disease damage of crops. The data acquisition forms are diversified, the parameter optimization quantity is large, the acquisition of large-range farmland parameters is realized, and a complete system is formed with remote sensing irrigation monitoring.

The invention discloses a multi-mechanism attribute-based encryption algorithm based on creditworthiness consensus and zero-knowledge proof, and relates to the field of blockchains. On the basis of traditional attribute-based encryption, creditworthiness evaluation is introduced, a decentralized trust model is established for evaluating and selecting trusted nodes, and the reliability of the algorithm is improved. According to the method and the system, the data can be shared and accessed in a credible environment, zero-knowledge proof is used for verifying the correctness and integrity of the data, specific data values or other sensitive information are not leaked, and the reliability of the data is improved by combining credibility consensus and the zero-knowledge proof technology. And an innovative and efficient solution is provided for data security sharing in a multi-mechanism environment.

The invention discloses a blockchain-based intelligent video monitoring data storage security sharing method for an iron and steel plant. The method comprises the following steps of: generating a public key PKS and a master key TMK based on an attribute encryption algorithm; distributing shares of the shared key S by constructing a matrix and defining a function, wherein the shares are distributed in different rows of the matrix; re-encrypting the data by using a hybrid proxy re-encryption algorithm based on a ciphertext strategy; secure data storage and access control are realized based on a secure storage mode of revocable attribute-based encryption RABE and a block chain; and in combination with a block chain technology, a consensus algorithm is adopted to verify transactions, so that the security and robustness of the system are enhanced. According to the invention, the utilization rate of the data is improved, the cooperation efficiency is enhanced, the security and transparency of the video monitoring data are improved, and the cost of information transmission and storage is reduced. The distributed storage and access control of the data are realized, the confidentiality and integrity of the cloud center data are ensured, and the security and efficiency of the steel plant video monitoring system in the storage and transmission process are enhanced.

The invention discloses an inter-satellite link secure transmission method based on a hybrid covert strategy, and belongs to the field of electronic communication. According to the invention, more randomness and friendly interference are introduced from three layers of a network layer, an MAC layer and a physical layer, and potential eavesdroppers are confused; in a network layer, a random routing unit is adopted, a plurality of routing paths meeting communication requirements are dynamically selected to replace a single optimal path, and a selection probability is distributed to each routing path according to routing metrics. In the MAC layer, the spectrum sharing technology is utilized to distribute less power for the information needing to be hidden, so that the signal detection probability is reduced, and the communication security is improved. In a physical layer, a rate division multiple access unit and a superposition coding technology are adopted, so that a plurality of users can efficiently share the same time-frequency resource at the same time. According to the invention, the communication efficiency can be improved, and the transmission randomness and complexity can be further improved, so that the communication security is enhanced.

The invention belongs to the field of traffic control systems, and relates to a vehicle driving track prediction system and method based on an improved graph neural network model, and the system comprises an FPGCN model and a Social LSTM model. The FPGCN model comprises a graph convolution layer, a Dropout layer, a differential pooling layer, a feature fusion module and a full connection layer; obtaining node feature vectors from graph structure data through a graph convolution layer, a Dropout layer and a differentiable pooling layer, then dividing and conquering node features, generating global features through a dense attention mechanism and principal component analysis, and generating local features through virtual filling and matrix calculation; after global features and local features are obtained, a feature fusion module is used for fusion, then the features are input into a full connection layer, and the influence score of a historical trajectory diagram on a current state diagram is calculated; integrating the preprocessed driving data and the data of which the score is calculated by the FPGCN model, and inputting the integrated data into a Social LSTM model; according to the system, while the model expression ability is maintained, the parameter quantity is remarkably reduced, the calculation logic is optimized, and the prediction accuracy is improved.

The invention provides a multi-scene fusion hyperactivity quantitative evaluation system, which is characterized in that motion data of a subject in different diagnosis scenes are acquired through a diagnosis scene data acquisition module, and motion behavior information of the subject is recorded; then, a single diagnosis scene feature extraction module is used for automatically extracting feature vectors corresponding to scene data of motion sensor time sequence data collected in the single task testing process by using a single scene representation network; using a multi-diagnosis scene efficient fusion module to take the feature vector of each scene data as the feature of a corresponding node on a graph, transmitting information among the feature representations of different scene data by adopting a graph neural network, integrating the scores of different options of a prediction scale, and optimizing a diagnosis process to reduce the task burden of a subject and improve the diagnosis accuracy of the subject. The objectivity and efficiency of symptom evaluation are remarkably improved, the accuracy and operability of ADHD clinical auxiliary diagnosis are effectively improved, and the method has wide clinical application prospects.

The invention discloses an LLM-based intelligent question and answer method and system for operation and maintenance knowledge, relates to the technical field of artificial intelligence, and provides an efficient knowledge management and update mechanism which can continuously absorb latest operation and maintenance knowledge and technical practice, and operation and maintenance personnel can upload new knowledge documents to the system through a data entry unit, so that the operation and maintenance knowledge can be quickly and conveniently answered. After verification of the data checking unit, the knowledge can be automatically partitioned and vectorized and supplemented into the knowledge base, the mechanism not only ensures timeliness and accuracy of the knowledge base, but also promotes accumulation and inheritance of the knowledge, new operation and maintenance personnel can quickly obtain necessary knowledge and skills through the system, and the operation and maintenance efficiency is improved. In addition, the system also supports the processing of multi-round dialogues and context relevance problems, so that the solution of complex problems is smoother and more efficient.

The invention discloses a multi-task reinforcement learning method based on an emergence hybrid expert model, and the method comprises the following steps: constructing a multi-task reinforcement learning network which comprises a plurality of feedforward neural networks, and connecting a routing network in front of each feedforward neural network; performing clustering analysis on each feedforward neural network, and equally dividing the feedforward neural network into a plurality of expert modules; training each routing network based on a strategy gradient descent method, and updating parameters of corresponding expert modules; selecting experiment tasks in the multi-task learning environment; and for each experiment task, activating a corresponding expert module based on the trained routing network to perform decoupling processing, and completing multi-task reinforcement learning. According to the method, an efficient and stable optimization strategy is provided for multi-task learning, and the problem of mutual interference between parameter sharing and tasks in multi-task learning is effectively relieved.

The invention discloses a swarm robot fault mutual diagnosis system and method based on cloud edge cooperation, and relates to the technical field of robot intelligent operation and maintenance. According to the invention, a cloud edge cooperation technology is adopted, efficient information sharing and cooperation are realized among a plurality of robots, and a distributed mutual monitoring and diagnosis mechanism is utilized, so that the states of the robots are monitored, the load capacity is reduced, and more functional modules can be carried; according to the Kalman filtering method based on multi-model adaptive estimation, the speed and displacement of the robot can be accurately estimated, a suspected fault robot can be more accurately found based on data consistency judgment deviation, and a cloud server further diagnoses according to historical data so that the judgment result can be more accurate. The cloud-side cooperation enables the maintenance personnel to find the fault more quickly and carry out maintenance in time, thereby guaranteeing the overall stability and reliability of the system.

The invention relates to the technical field of communication equipment, and provides a heat dissipation device and method based on idle heat dissipation resource sharing in a case, and the device comprises heat dissipation fins equipped for a plurality of service ports, heat conduction graphite flakes and an automatic fan system. The cooling fins are lengthened and connected through the heat conduction graphite flakes to form an annular heat dissipation network, free conduction of heat among all service ports is achieved, and idle heat dissipation resources are effectively utilized and shared. The automatic fan system automatically adjusts the wind speed according to the highest temperature in the case to ensure that the temperature is appropriate. The temperature sensor monitors the temperature of each service port in real time, and provides accurate feedback for a fan system and a control system. Besides, the device also has comprehensive monitoring and optimizing functions, continuously monitors the temperature distribution in the case, the running state of the fan and the noise level, and performs optimization and adjustment as required. According to the heat dissipation device and method, the heat dissipation efficiency is improved, the energy consumption and the noise are reduced, and the stable operation and the optimal heat dissipation effect of the case are ensured.

The invention discloses an enterprise information sharing management method, system and device and a storage medium, which are convenient for managing enterprise data, can promote cooperation and communication among departments in an enterprise, and can improve the utilization efficiency and value of data. The method comprises the following steps: acquiring department data uploaded by each department, and judging whether the format of the department data uploaded by each department accords with an uploading format or not through a regular expression; when it is determined that the format of the department data accords with an uploading format, a unique identification identifier is automatically generated according to the department data, and the identification identifier comprises a timestamp and department information corresponding to the department data; classifying the department data according to department information to obtain classified data; uploading the classified data to an information management system; and when application information of target data is received, judging whether the target data has the access permission or not according to the application information.

The invention belongs to the technical field of edge computing power scheduling, and relates to an edge computing power scheduling method, device and equipment based on time delay perception optimization. The method comprises the following steps: acquiring an edge computing environment, and constructing a multi-edge service provider computing power sharing framework; according to a multi-edge service provider computing power sharing framework, the service time delay is minimized as a target, and the computing task load undertaken by edge service providers, the scheduling decision of the edge service providers, the scheduling matrix of the edge service providers and the computing task load processed by the edge service providers are constrained. Establishing an optimization problem of edge computing power task scheduling; solving an optimization problem of edge computing power task scheduling to obtain a scheduling matrix of multiple edge service providers; and according to the scheduling matrix of the multi-edge service providers, performing edge computing power scheduling of the multi-edge service providers. By adopting the method and the device, efficient computing power resource sharing and efficient edge computing power task scheduling can be realized.

The invention discloses a park energy sharing and distributed scheduling method based on a cooperative game, and the method comprises the steps: 1), constructing an energy storage model, an electric vehicle model, an air conditioning system model and a photovoltaic model, and forming a park equipment model; 2) in combination with the park equipment model, constructing a park energy sharing and distributed scheduling model based on a cooperative game; and 3) solving the park energy sharing and distributed scheduling model by adopting an ADMM algorithm to obtain a park energy sharing and distributed scheduling strategy. According to the invention, a park energy sharing and distributed scheduling model based on a cooperative game is established, and an ADMM algorithm is adopted to solve and obtain a park energy sharing and distributed scheduling strategy, so that efficient scheduling and cooperative operation of resources are realized, nearby consumption of distributed energy is facilitated, the advantage of resource sharing is exerted, and the energy utilization efficiency is improved. The operation stability and reliability of the park are improved, information interaction in the scheduling process is reduced, and privacy protection is achieved.

The invention relates to a power network system abnormal access detection method and device and a medium, and the method comprises the following steps: obtaining power network system data, inputting a power network system abnormal access detection vertical model, and outputting an abnormal access detection result; the power network system abnormal access detection vertical class model is obtained through the following steps that S1, power network security corpus data are obtained and preprocessed, and a pre-training data set is constructed; s2, pre-training the big language base model to obtain a power network security big language model; s3, performing secondary screening on the pre-training data set based on an abnormal access detection task of the power network system, and constructing a supervision fine tuning data set in combination with expert opinion data; and S4, performing fine tuning on the power network security big language model to obtain a power network system abnormal access detection vertical class model. Compared with the prior art, the capability of accurately detecting various abnormal access behaviors in a novel network system with a large amount of multi-modal data can be improved.

The invention provides a full life cycle data management system and method based on a train configuration. The method comprises the following steps: establishing a train function configuration; wherein the train function configuration comprises a system, a subsystem, a function group and a function part of a train; building a design structure tree, wherein the design structure tree comprises design data; associating design data with the train function configuration to generate a train physical configuration; and associating the train physical configuration with the service management platform in the train full life cycle, and realizing the circulation of the train full life cycle data in different service management platforms on the basis of the train physical configuration. According to the invention, a configuration management method is introduced, the inherent architecture of a service system is not changed, the influence on the service flow of each interface system is very small, and through the active and automatic association of the configuration, the communication of full-life-cycle data, upward tracing to a supplier production process and downward extending to a customer operation and maintenance management process are realized; and the full-life-cycle data of the industrial chain is integrated.

The invention discloses a theme style sharing method and device and a server, and relates to the technical field of computers.The method comprises the steps that a first terminal adjusts configuration parameters of a local artificial intelligence model in response to personalized configuration operation on theme styles, and style configuration parameters are obtained; wherein the artificial intelligence model generates a target digital asset of a personalized theme style corresponding to the personalized configuration operation under the condition that the artificial intelligence model runs based on the style configuration parameters; performing structured processing on the style configuration parameters to generate a style configuration file; and sharing the style configuration file to a second terminal to enable the second terminal to restore the style configuration parameters based on the style configuration file, and generating a target digital asset based on a local artificial intelligence model and the style configuration parameters of the second terminal to reproduce the personalized theme style. According to the method and the device, the user can be supported to perform personalized creation on the theme style and perform efficient sharing on the creation result, and meanwhile, the shared theme style can be ensured to be accurately reproduced at the receiving end.

The invention discloses an industrial data analysis method and system based on a dynamic form, and relates to the technical field of data analysis, when a plurality of systems share data, the systems calculate the similarity Xsd of data items through a repeatability analysis module, and screen out repeated data sets based on the similarity, so that the systems can effectively identify and eliminate redundant data, and the data processing efficiency is improved. And unnecessary repeated acquisition and storage are avoided, so that data storage is optimized, storage resources are saved, and the burden of redundant data is reduced. Through a data usage analysis module, the system can identify features of each data item in different contexts within a duplicated data set. Through data context vectorization and analysis of the use difference degree Cyd in different contexts, the system can further judge which data items have different meanings in different systems or service scenes, so that non-redundant data which should not be regarded as redundancy is screened out, and the analysis ensures that the system performs judgment not only based on the data similarity.

The present invention relates to the field of data analysis and processing technology, and in particular to an intelligent warehousing system and method based on supply chain management. First, the original data is obtained and transformed to obtain market trend data and extract market trend data features, and a time series decomposition method is introduced to perform trend prediction through a deep learning model. Seasonal adjustments are performed through high-order Fourier transforms, and an integrated learning model is introduced to generate predictions for future inventory demand; secondly, a real-time data processing framework and an adaptive data analysis model based on edge computing are introduced, compatibility and conflict resolution mechanisms are designed, modular design and plug-in architecture are adopted, and a comprehensive maintenance tool is introduced. This solves the problem that inventory prediction in the prior art relies on traditional prediction methods, cannot accurately process complex market data, resulting in inaccurate inventory predictions and excess or shortages of inventory; and limits the ability to extract useful information from complex data, resulting in poor system performance.

The invention provides a cross-domain cooperative network exception information sharing method and device, equipment and a medium, relates to the technical field of network security, and is used for solving the technical problem that in the prior art, a network exception detection method mostly depends on data in a single domain and is difficult to adapt to security requirements in a cross-domain network environment. The method comprises the following steps: constructing a trust framework based on a plurality of network domains, the trust framework being configured to enable information sharing among the plurality of network domains; unifying a data exchange protocol among the network domains in the trust framework; in response to the abnormality of at least one network domain in the plurality of network domains, performing root cause analysis on the abnormality to obtain an analysis result; and the analysis result is shared to the plurality of network domains through the trust framework, so that the plurality of network domains can respond in real time according to the analysis result.

The invention discloses an early warning method based on a block chain and a sub-control server, and relates to the technical field of production monitoring of each workshop of an automobile, and the method comprises the steps: S1, monitoring the production state of each process section of each workshop based on a PLC, if the production state of a certain process section is abnormal, generating fault information of a fault process section of a target monitoring workshop, the sub-control server reports the fault process section; s2, the sub-control server of the fault process section analyzes the fault information and generates a fault information block of the fault process section; s3, the information block of the fault process section is broadcasted and linked, a new block is generated through verification of other nodes, and the nodes of the workshops are synchronously updated; and S4, each node of each workshop analyzes and obtains corresponding production information according to the received information block of the fault process section, carries out early warning and adjusts a production plan. According to the invention, transparency and tampering resistance of the block chain technology are combined, and an efficient mode is provided to enhance the automobile production monitoring and early warning system.

The present invention discloses a method and system for managing cross-regional parking information in a smart city, and relates to the technical field of parking management. The method includes: reading parking information elements within a preset area range, and constructing a cross-domain information architecture; traversing the cross-domain information architecture, and constructing an information consensus model; traversing the cross-domain information architecture, and coding and labeling data nodes, and constructing a dual-drive management module; authorizing parking rights based on the user end, and generating parking services; taking the parking rights as constraints, making cross-domain judgments on the parking services; if it is a cross-domain parking service, performing consensus trigger conversion for policy execution; obtaining service execution information, and synchronously updating and storing parking information based on an intercommunication platform. The present invention solves the technical problems in the prior art that parking information in different regions is dispersed, difficult to uniformly manage and integrate, and the user parking experience is poor, and achieves the technical effect of realizing efficient integration and sharing of parking information and improving the user parking experience.

The invention discloses an operation cost pre-estimation method and system based on machine learning, and relates to the technical field of enterprise operation management, the operation cost pre-estimation method based on machine learning adopts a distributed computing architecture, improves the efficiency of data processing and model training, divides operation data to a plurality of computing nodes for parallel processing, and improves the efficiency of data processing and model training. According to the method, the time period of cost estimation can be remarkably shortened, an enterprise can obtain an accurate cost estimation result in time, market changes are quickly responded, an operation strategy is adjusted in time, the distributed computing architecture further has good expandability, and the cost estimation efficiency is improved along with increase of enterprise business and increase of data volume. The computing nodes can be conveniently added to meet new requirements, the hardware upgrading cost and the system maintenance complexity of an enterprise are effectively reduced, and powerful technical support is provided for long-term development of the enterprise.

Provided is an information processing apparatus that allows a plurality of systems to efficiently share information regarding an alert. An information processing apparatus includes an ID assigning computer that assigns a log ID to collected log data of a device, a first processing computer that stores the log data and the log ID and detects an abnormality based on the log data, and a second processing computer that stores the log data and the log ID, detects an abnormality based on the log data, and transmits the log ID of the log data in which the abnormality is detected and a detection result of the abnormality to the first processing computer. The first processing computer stores the log data stored in the first processing computer and the received detection result in association with each other based on the log ID received from the second processing computer.

This invention provides a tank container for intermodal transport, comprising: a frame and a tank body horizontally installed inside the frame; the frame includes two end frames respectively connected to the longitudinal ends of the tank body; the frame also includes two sets of support seats and multiple reinforcing members evenly disposed at the bottom of the tank body; the two sets of support seats are longitudinally spaced at both ends of the tank body, each set having two support seats spaced laterally along the tank body; the support seat is an internally hollow shell, the support seat having an opening communicating with the interior of the shell, the opening facing upwards, and the support seat is located at the periphery of the opening and is connected and fixed to the bottom of the tank body; each support seat extends longitudinally, and one end of the support seat along the longitudinal direction is connected to the end frame; the reinforcing members are received inside the support seat through the opening, and the reinforcing members are respectively connected and fixed to the inner wall of the support seat and the bottom of the tank body, thereby achieving support and reinforcement of the support seat, increasing its impact resistance, and enabling the support seat, as a load transfer area, to effectively transfer and distribute the load of the tank body.

The application discloses a kind of robot post-disaster rescue collaboration methods based on deep learning and large language model, by introducing " reflection mechanism " in natural language processing, realize task semantic analysis, dynamic optimization and adaptive scheduling.Large language model generates task semantic graph and combines reinforcement learning to adjust priority in real time, form closed-loop task reflection and optimization process, greatly improve the flexibility and adaptability of robot in complex environment.Design " distributed perception bridge " module, through attention mechanism dynamic screening and sharing key perception data, realize the efficient cooperation of heterogeneous robot, even in the condition of unstable communication still can guarantee rescue efficiency.Based on the reinforcement learning model driven by environmental characteristics, real-time perception of physical properties such as fragment density and slope, optimize behavior strategy, improve the generalization ability and robustness of model.Through " force-semantic coupling " motion planning, dynamically adjust the operation path and force of manipulator, improve the success rate of complex task.