70 results about "Exponential weighting" patented technology

An AI-driven financial planning system for real-time market adjustment, consisting of: a neural inference coprocessor configured to execute deep financial forecasting models, including recurrent neural networks and attention-based encoders, on the device, and wherein the processor dynamically updates portfolio parameters in response to market signals exhibiting volatility differences above a statistical threshold calculated using an exponentially weighted moving standard deviation; a financial data acquisition module configured to continuously receive and analyze heterogeneous data streams, including market indices, interest rates, stock and bond price fluctuations, economic indicators, regulatory updates, and financial news sentiment feeds; a behavioral analytics engine configured to create a dynamically evolving user-specific financial behavior profile based on real-time analysis of transaction history, income-expenditure cycles, psychometric test results, and temporal lifestyle patterns using supervised and unsupervised machine learning algorithms; A goal optimization module configured to transform high-level, user-defined financial goals into quantitatively tracked multi-level goals. It uses a reinforcement learning framework that predicts optimal asset allocations across multiple time horizons. a real-time strategy simulation engine configured to perform Monte Carlo simulations and deep Q-learning-based assessments to simulate the resilience of proposed financial strategies under different macroeconomic regimes and trigger redistribution events based on predefined confidence thresholds; a compliance-aware execution interface configured to interact with financial institutions through encrypted API channels, ensuring policy enforcement using a smart contract validator and a hardware-enabled secure transaction signing unit; and a recommendation display unit configured to render dynamic dashboards for visualizing investments, reallocation warnings, confidence intervals, and sensitivity sliders, and where user interaction with the unit flows back into the behavioral model for real-time learning.

The invention relates to the technical field of energy system control, and discloses a distributed cooperative fault-tolerant control method for a multi-energy-station cooling and heating system, and the method comprises the steps: enabling each energy station to measure and calculate a residual error based on a physical model and a sensor, forming a normalized health index, and carrying out the neighborhood broadcasting; carrying out robust anomaly judgment by utilizing a neighborhood median and a median absolute deviation, and isolating an abnormal site; cost coefficients are automatically generated for available sites according to equipment maneuverability, and control redistribution of minimum disturbance is solved and implemented through distributed consistency optimization; an exponential weighting updating mechanism of a residual error sequence is adopted to carry out online self-adaption on a noise baseline and a judgment threshold value so as to realize closed-loop self-calibration; when sensor abnormity, equipment errors or individual station faults occur in the multiple energy stations which work cooperatively, distributed rapid identification is achieved, faults are isolated, energy supply is redistributed with minimum disturbance, system control requirements are met, and equipment safety constraints are kept.

The invention discloses a multi-source navigation data fusion method for an unmanned loader, which comprises the following steps: data acquisition: acquiring environmental perception data of satellite navigation, an inertial measurement unit IMU, a wheel type odometer and a laser radar and camera in real time; performing space-time synchronization preprocessing, realizing multi-source data time synchronization, unifying environment sensing data to a body coordinate system, eliminating abnormal data, and complementing missing data; carrying out dynamic weight calculation, establishing an error model of each sensor, and adjusting a fusion weight by using an error reciprocal exponential weighting method; layering fusion is carried out, a satellite and an IMU are fused through bottom-layer extended Kalman filtering (EKF), a laser radar and a high-precision map are fused through middle-layer ICP, a middle-layer result and a wheel type odometer are integrated through high-layer federated filtering, and high-precision fusion is achieved; and outputting and optimizing a result, outputting navigation data, performing closed-loop optimization, monitoring the health degree of the sensor, and executing redundancy switching when a fault occurs. The system comprises a corresponding processing unit, and a storage medium stores a program for executing the method.

The invention relates to a tea-picking robot autonomous navigation method based on tea ridge composite boundary identification and driving area adjustment, and belongs to the technical field of robot navigation. The method comprises the steps that tea ridge growth state information is evaluated, and a global path planning strategy is made; a root-ground intersection ground navigation datum line and lateral boundary constraints are established, the ground navigation datum line, the lateral boundary constraints and tea tree growth density and distribution characteristics are integrated, a tea ridge three-dimensional drivable road network model is established, and key geometric characteristics of the model are analyzed to establish an adaptive risk assessment system. Dividing the drivable area into areas with different risk levels; and establishing a multi-objective optimization function in combination with the risk region distribution diagram, fusing the multi-objective optimization function through an exponential weighted fusion mechanism, and generating an optimal navigation path under the risk gradient constraint. Guiding of intelligent navigation of the tea garden is achieved, the tea leaf picking efficiency and quality are improved, and technical support is provided for mechanical and intelligent picking of the tea garden.

The invention relates to the technical field of water, wind and light storage combination, in particular to a wind, light and water storage combination robust tracking method and device and a medium, which can uniformly represent the output uncertainty of extreme renewable energy sources and quickly map a safe feasible region in a high-dimensional scheduling space. And safe and economic cooperative control of the hydroelectric generating set, the reservoir water level and the energy storage resources within the resolution ratio of 15 minutes is realized. Firstly, unified normalization and sliding window time embedding are performed on wind speed, irradiation and load sequences in a data layer, and a dynamic error envelope is constructed through exponential weighting of a high-tail sample, so that a measurable extreme interval is directly obtained in an original observation space. And then introducing a physical climbing limit and a unit inertial constraint, sequentially expanding a disturbance scene tree on nodes at two ends of the error envelope, carrying out error envelope-physical mapping coupling modeling, enabling the system to keep time-space correlation, and directly incorporating unit dynamic characteristics in a scene generation stage.

The invention discloses an intelligent evaluation system and method for knowledge base question and answer application, and relates to the technical field of artificial intelligence, natural language processing and multi-agent collaborative systems. The system comprises a data preprocessing module, a data synthesis module, a data screening module, a data scoring module and a self-adaptive weight adjustment module. The data preprocessing module carries out data preprocessing on the enterprise original document and generates a data knowledge base; the data synthesis module adopts a large language model to perform repeated question and answer on each paragraph of the data knowledge base to generate question and answer pairs; the data screening module screens out high-quality samples through a screening agent collaborative screening MACA framework mechanism; the data scoring module performs question and answer scoring through a scoring agent collaborative scoring MACA framework mechanism; and the adaptive weight adjustment module dynamically updates the weight of each screening scoring dimension by adopting an exponential weighted moving average algorithm.

The invention provides a multi-modal feature fusion identification method and device for partial discharge of a high-voltage cable, and relates to the technical field of multi-modal feature fusion of partial discharge of the high-voltage cable. Multi-modal time sequence data of three-phase current, temperature and voltage to ground of the high-voltage cable are collected and subjected to preprocessing and time alignment to form a multi-dimensional input matrix; a physical relation matrix based on cable physical characteristics is constructed, physical constraints are injected into data through feature modulation, fused feature representation is generated, features are input to an LSTM auto-encoder trained by normal data, a sequence is reconstructed, a reconstruction error is calculated, finally, an exponential weighting algorithm is adopted to accumulate an error time sequence to obtain a comprehensive abnormal score, and the comprehensive abnormal score is obtained. And through comparison with a dynamic threshold value, accurate identification and partial discharge alarm of the insulation state abnormity are realized.

The invention relates to a YOLOv4-based optimization method for rapid identification and detection of a sowthistle obstacle, and the method comprises the steps: initializing an anchor frame through employing a K-means algorithm, generating an anchor frame with higher scale adaptability, and enabling a model to be more suitable for the detection of slender sowthistle; aiming at network structure optimization of network scene particularity, cutting far all-grass of Chinese ixeris and near small all-grass of Chinese ixeris, and only stably detecting a close shot existing in each frame of image; in the optimization of a pooling mode, pooling is carried out by adopting an exponential weighted average filtering mode so as to reserve all useful information as much as possible. According to the method, the YOLOv4 is adopted to accurately detect the all-grass of Chinese Ixeris, optimization is carried out in three aspects of priori frame reclustering, a pooling mode and a network structure, and the optimized YOLOv4 can accelerate a target detection process, so that the method is suitable for real-time application, the use of calculation and memory resources is reduced, the method is suitable for an embedded system or equipment with resource limitation, and the detection efficiency of the all-grass of Chinese Ixeris is improved. The accuracy of target detection is not damaged as much as possible, and high-quality recognition of the target is kept.

The invention discloses a motor and a control method and device thereof, a storage medium and a computer program product, and the method comprises the steps: injecting a pulsating voltage with a preset frequency into a d-axis part outputted by a PI controller in a current loop at a zero-speed starting stage and a low-speed operation stage of the motor; in the whole stage of the motor, three-phase current of the motor is obtained; in the whole stage of the motor, in a current loop and a speed loop, according to the three-phase current of the motor, a full-order observation module, a phase locking module and a switching function module are utilized to obtain a rotor position estimation value and a rotating speed estimation value of the motor; and in the whole stage of the motor, starting and running of the motor are controlled according to the rotor position estimation value and the rotating speed estimation value of the motor. According to the scheme, the improved high-frequency injection algorithm is combined with the full-order observer based on the back electromotive force, and the exponential weighting function is adopted to smoothly switch the position and the speed of the motor rotor estimated by the high-frequency injection algorithm and the full-order observer, so that the complexity and the cost are reduced, and the reliability is improved.

The invention discloses an elevator prediction scheduling method and system based on user habit self-learning, and the method comprises the steps: carrying out the statistical analysis based on historical elevator taking event data, and judging whether a floor combination meeting a preset condition exists or not; combining all the independent floors with the floors meeting the preset conditions to serve as a self-learning unit; for each self-learning unit, updating the elevator calling probability of the respective learning unit by using an exponential weighted average algorithm; when the elevator does not have the real-time task, all elevator calling probabilities corresponding to the current time window are inquired, and the self-learning units meeting the triggering condition are extracted as a target candidate set; calculating an optimal pre-stop layer by utilizing a mathematical model based on the target candidate set, and calculating a comprehensive probability corresponding to the optimal pre-stop layer; and calculating an income evaluation function based on the comprehensive probability, and generating an instruction to drive the elevator to run to an optimal pre-stop layer to enter a prediction waiting state when an obtained value meets a condition.

The application discloses a kind of fusion and compression method of multi-source trajectory information, it is related to data processing field.The application uses the thought of smoothing analysis, by adaptive calculation angle, speed error threshold, each positioning terminal trajectory is denoised in real time;Real-time calculation window time each trajectory data quality;According to the weight of each trajectory of trajectory data quality adaptive calculation, improved exponential weighted average algorithm is proposed, the weighted average value of longitude and latitude of multi-source trajectory point is calculated as the fused trajectory coordinates;The roughness of fused trajectory in sliding time window is calculated, and the sub-block size of GeoHash algorithm is dynamically adjusted by integrating trajectory roughness, to realize efficient real-time compression under the premise of maintaining high compression accuracy.The application can realize adaptive fusion and compression of multi-source trajectory by comprehensive data quality, with the advantages of reasonable fusion weight and high real-time compression efficiency.

The present disclosure provides a hardware resource allocation method, device and equipment of an electronic device, which preliminarily identifies hardware power consumption mutation by sampling the hardware power consumption data of the electronic device in real time and determining the fluctuation index corresponding to the current sampling data. Further, the exponential weighted value of the current sampling data is determined, and when the current exponential weighted value and the exponential weighted value when the last fluctuation index meets the set condition, the mutation is determined as an effective mutation. Furthermore, when the fluctuation index and the exponential weighted value meet the condition at the same time, the future hardware power consumption data is further predicted, and the hardware resource demand information of the electronic device in the next target period is determined in combination with the fluctuation index. The fluctuation of the hardware power consumption data can cover the user's fine-grained operation, and by predicting the future hardware resource demand information, the resources can be prepared in advance, avoiding the situation of insufficient or lagging resources during scene switching, and ensuring smooth user experience.

This filtering method comprises steps in which: a Kalman gain is calculated; a system calculates an estimated value on the basis of the Kalman gain; a predicted value and a measured value of an error covariance are compared and analyzed to calculate the error covariance; an exponentially weighted average and an exponentially weighted covariance for the estimated value are calculated; and the exponentially weighted average and the exponentially weighted covariance of the calculated estimated value are reflected in the subsequent Kalman gain calculation. Therefore, performance superior to that of a conventional Kalman filter can be achieved in non-Gaussian noise conditions.

The application discloses an intelligent evaluation system and method for a knowledge base question answering application, and relates to the technical fields of artificial intelligence, natural language processing and multi-agent collaborative system. The system comprises a data preprocessing module, a data synthesis module, a data screening module, a data scoring module and an adaptive weight adjustment module. The data preprocessing module performs data preprocessing on enterprise original documents and generates a data knowledge base. The data synthesis module repeatedly asks and answers each paragraph of the data knowledge base by using a large language model to generate a question and answer pair. The data screening module screens out high-quality samples through screening agent collaboration and a MACA framework mechanism. The data scoring module scores the question and answer pair through scoring agent collaboration and a MACA framework mechanism. The adaptive weight adjustment module dynamically updates the weight of each screening and scoring dimension by using an exponentially weighted moving average algorithm.

The application discloses a method for removing and realizing strong vibration interference in a force signal of a polishing robot, two filter algorithms are connected in series to form a double Kalman filter, a first filter realizes filtering of Gaussian white noise and high-frequency vibration spectrum peak group noise; a single parameter is introduced, a gradually time-varying noise variance is designed in an exponential weighting mode to describe the colored noise characteristics, and an improved Kalman filter is designed to form a second filter to realize filtering and elimination of the colored noise. The application realizes effective filtering and elimination of complex noise composed of Gaussian white noise, vibration noise and colored noise superimposed in force measurement information in robot force control polishing.

The invention discloses a Bernoulli filter-based single target detection and DOA tracking method and system, and relates to the field of target tracking. The method solves the problems that a traditional tracking algorithm after detection cannot jointly detect the target when the underwater target repeatedly enters under the conditions of low signal-to-noise ratio and small snapshot, and the tracking effect is unstable. The method comprises the steps of constructing a Bernoulli random finite set model, defining a single target state as a dynamic description target state or a dynamic description target state, predicting the target existence probability by using a Bernoulli Markov process, and generating a prediction state distribution particle set composed of new particles and survival particles through particle filtering. A hydrophone array signal is decomposed into a noise subspace observation model and a signal subspace observation model, a generalized likelihood function after exponential weighting is used for updating a target posteriori existence probability and a particle weight, and an equal-weight particle set is generated through resampling. And judging whether the target exists or not according to the single-target posterior existence probability, thereby realizing single-target DOA joint detection and tracking.

The invention relates to a dynamic capacity expansion and contraction method and device for a database thread pool. The method comprises the following steps: periodically collecting indexes such as session response and processing delay, request queue length and active thread number of a database system; smoothing processing is carried out on the collection indexes to eliminate the influence of instantaneous data fluctuation; and judging the current load state (overload, normal or idle) of the system according to the indexes, if the current load state is the overload state or the idle state, executing the step-by-step thread pool capacity expansion and contraction operation according to the system resource occupation condition, and returning to form closed-loop feedback regulation after the execution is finished. The method has an adaptive dynamic adjustment capability, and accurately senses the system state through an exponential weighted average algorithm; intelligent decision-making and stepping capacity expansion and contraction strategies are adopted, and efficient adjustment of'changing from quantity to demand 'is realized in combination with resource occupancy detection; and finally, self-adaptive optimization of the database thread pool is realized, the resource utilization rate is improved while stable performance is ensured, the manual intervention cost is reduced, and an efficient technical solution is provided for a high-concurrency scene.

The invention relates to the technical field of control systems, and discloses a real-time calculation and analysis method and system for equipment comprehensive efficiency (OEE), and the method comprises the steps: collecting the operation data of equipment, the operation time of the equipment, the planned operation time and the theoretical cycle time in real time; performing exponential weighted fusion of multi-source data on the equipment operation time, the planned operation time and the theoretical cycle time to obtain a time starting rate; based on the time starting rate and the real-time performance starting rate of the equipment, performing dynamic coupling analysis on the operation efficiency index of the equipment to obtain a comprehensive efficiency OEE value; performing real-time state analysis on the comprehensive efficiency OEE value to obtain an efficiency analysis result of the equipment; based on the efficiency analysis result, efficiency parameter optimization is carried out on the equipment, and optimized operation configuration is obtained; based on the optimized operation configuration, performing effect evaluation on the operation parameters of the equipment to obtain a final efficiency optimization report; according to the invention, the accuracy of real-time calculation and analysis of the equipment comprehensive efficiency OEE can be improved.

The invention discloses an in-situ monitoring data anomaly detection method and device based on trend decomposition and state modeling. The method comprises the following steps: preprocessing an ontology sequence and an environment sequence acquired in an in-situ monitoring area; decomposing the two types of sequences into a long-term trend term, a seasonal term and a residual term, normalizing, fragmenting and splicing into a fragment set; a fragment set is extracted in batches, and value, position and time coding and fusion are carried out on each component fragment; inputting each coding segment into a spatial state modeling module and an attention mechanism module, fusing results to obtain a predicted long-term trend, seasons and residual errors, and carrying out reverse normalization and splicing fusion to obtain a final prediction result; and processing a to-be-predicted ontology sequence and a corresponding environment sequence according to the previous steps, reasoning by using the trained model to obtain ontology prediction fragments, splicing the ontology prediction fragments into a continuous prediction sequence, calculating a prediction residual error, and performing bilateral exponentially weighted moving average anomaly detection. According to the invention, the accuracy and stability of anomaly detection can be effectively improved, and the timeliness and reliability of early warning are improved.

The invention discloses a trace element full-automatic detection system and method based on reinforcement learning. The method comprises the following steps: extracting target area and channel mark information by using image analysis, constructing a detection task atlas structural body, performing multi-round path simulation and return calculation in combination with a Monte Carlo tree structure, and generating a detection strategy path set. And detecting an action, selecting an action probability distribution tensor constructed based on an Exp3 algorithm, and realizing dynamic scheduling according to an exponential weighting weight. In the detection process, an actual result is collected, deviation calculation is carried out on the actual result and path simulation return, a feedback error vector is generated, a Monte Carlo tree structure and action probability distribution are updated, a strategy convergence state is judged, whether a strategy path set is reconstructed or not is controlled, and self-adaptive optimization of a detection strategy is achieved. According to the invention, self-adaptive scheduling and accurate control of the trace element detection process are realized.

The invention discloses a face velocity real-time monitoring method and system based on a ventilation cabinet, and relates to the field of anomaly detection, and the method comprises the steps: synchronously collecting a ventilation valve opening value, a pipeline internal pressure value and a face velocity value; step division is carried out according to the valve opening adjustment values, the adjustment values are calculated, hysteresis of pressure and wind speed response is considered, data are classified through a clustering algorithm, and step data under all the adjustment values are obtained; calculating an abnormal coefficient of each valve adjustment value based on time interval and deviation statistics of step data, and using an exponential weighting method according to the coefficient in combination with data recency and deviation degree; normalizing the abnormal coefficient according to the occurrence frequency of the opening degrees of the valves to obtain a final abnormal value of each opening degree; and judging whether the ventilation cabinet is normal or not by comparing the final abnormal value with a preset threshold value. The method effectively improves the monitoring accuracy, adapts to the accumulation characteristic of chronic faults, and ensures the safe operation of a ventilation system.

The invention relates to an RSSI-based indoor positioning method and device, a storage medium and electronic equipment, and relates to the technical field of indoor positioning, and the method comprises the steps: obtaining an RSSI value of at least one target signal received by a to-be-positioned target in an indoor region; according to the sequence of the RSSI values from large to small, selecting a preset number of signal transmitting nodes from the signal transmitting nodes, and determining the selected signal transmitting nodes as participating positioning nodes; according to the RSSI value of each participating positioning node, determining a target distance between the corresponding participating positioning node and the to-be-positioned target; determining an initial positioning coordinate corresponding to the to-be-positioned target according to the initial weighting coefficient and the coordinate information corresponding to each participating positioning node; and determining a final positioning coordinate corresponding to the target to be positioned through a preset weighted centroid positioning algorithm according to the exponential weighting coefficient and the coordinate information corresponding to each participating positioning node. The method has the effect of improving the indoor positioning precision.

The invention belongs to the technical field of streaming media big data, and particularly relates to a video prediction caching strategy based on a Markov correction model, which comprises the following steps: S1, extracting a user access rule according to a user access record, namely prior data, and obtaining an initial state transition matrix and a user access initial probability; s2, correcting the state transition matrix by using an exponential weighted average model, and adding an old state transition matrix into a new state transition matrix in a weighted summation mode; s3, the state transition matrix of the prediction segment is obtained based on iterative calculation of the state transition matrix, and then the accessed probability of the video segment at each time point of the prediction segment is calculated; and S4, selecting a corresponding video segment for caching based on the size of the accessed probability. According to the method, the access times and frequencies of the recent video segments can be fully considered, the popularity of the newly online video segments can be fully considered, and the influence of premature historical data on the prediction accuracy of the system on the recent video popularity is avoided.

The invention relates to the technical field of high-voltage relay life prediction, and discloses a high-voltage relay residual life prediction method based on an implicit non-linear scale transformation wiener process, and the method comprises the steps: carrying out the smoothing of the arcing time degradation data of a high-voltage relay through employing a moving average method based on the exponential weighting reinforcement recent influence; dividing different degradation stages of the high-voltage relay by using an important point clustering algorithm, extracting later-stage degradation data of arcing time, extracting a low-frequency intrinsic mode function component reflecting degradation characteristics of the high-voltage relay in the later-stage degradation data of the high-voltage relay through variational mode decomposition, and verifying feasibility of a wiener process; meanwhile, a degradation model is established for the later degradation stage of the high-voltage relay by using an implicit nonlinear scale transformation wiener process, model parameters are estimated, the influence of system noise is considered at the same time, and a drift coefficient is updated online through a Bayesian recursion method; and finally, predicting the residual life of the high-voltage relay based on the wiener process model in combination with the failure threshold value of the high-voltage relay, and evaluating the reliability.

The invention provides a traffic flow data intelligent monitoring method and system for a highway toll station, and belongs to the technical field of data monitoring. Comprising the following steps: preprocessing acquired traffic flow information into traffic flow data to be monitored, storing the traffic flow data to a corresponding database, and calculating current traffic flow data of each lane by adopting a time division method to determine fluctuation data of each lane; the method comprises the following steps: modeling a highway toll station and a surrounding environment based on a digital twin platform, and displaying fluctuation data of each lane; meanwhile, analyzing the fluctuation data of each lane by adopting an exponential weighted decline and skyline algorithm; and if the fluctuation data in any lane is abnormal, executing an alarm operation. According to the invention, accurate abnormity monitoring can be carried out for traffic flow characteristics with different date properties, potential peaks and abnormal fluctuations can be found in time, and a scientific decision basis is provided for lane management, traffic scheduling and safety guarantee of a highway toll station.

The invention provides a civil aviation operation-oriented airspace probability modeling and risk quantification method, which uses ADS-B data to construct layered space-time probability density, adopts an anisotropic kernel which is widened along the course and tightened in the transverse direction, and realizes short-time smoothing by using a time kernel. Inverse detection probability correction is introduced into the sample weight to reduce the coverage deviation. The reference density is subjected to multi-day exponential weighting generation according to hour data so as to describe day and night laws and workday characteristics. Under an information geometry framework, a relative entropy item, a Fisher information item, a capacity penalty item and a conflict kernel item are combined into a unified risk functional, and calculation of density evolution and a congestion dissipation direction is realized by a Wasserstein gradient flow. The system is composed of a data preprocessing module, a detection probability correction module, a layered anisotropic space-time kernel density estimation module, a reference density generation module, a risk functional and gradient flow module and an increment updating and index output module. The technology has unified modeling and clear interpretable and engineering landing characteristics, and can be used for situation analysis and collaborative decision support of air routes, sectors and terminal areas.

The application discloses a kind of based on multi-modal data fusion's intelligent early warning system and method of accompanying, belong to wisdom medical treatment and artificial intelligence technical field, this system includes multi-modal data acquisition and pre-processing module, dynamic weight fusion module, intelligent risk assessment module and multidisciplinary collaborative intervention module, through hardware synchronization and software interpolation, the time-space alignment of multi-source sensor data is realized, the dynamic weight fusion mechanism combining multi-head self-attention and evidence theory is used to process modal conflict, risk dynamic assessment is carried out using sliding window exponential weighting and LSTM time series modeling, and based on medical knowledge graph and improved hungarian algorithm, the intelligent scheduling and closed-loop feedback optimization of nursing resource are realized.

The invention relates to the field of artificial intelligence and natural language processing, and particularly provides a language model dynamic activation guide control method and system based on semantic feedback. The method comprises the following steps: S1, extracting intermediate layer activation in a language model generation process, and obtaining an interpretable semantic state through a sparse auto-encoder; s2, calculating the real-time difference between the current semantic state and the target semantic state; s3, dynamically calculating the current optimal guiding strength by adopting exponential weighting accumulation based on the historical semantic difference sequence; and S4, injecting the activation guide vector corresponding to the optimal guide intensity into the language model so as to adaptively control the output content of the language model in real time. According to the method, manual parameter adjustment is not needed, the problem of insufficient or excessive control in static guidance can be effectively avoided, the semantic consistency and quality of the generated text are remarkably improved, and the method can be widely applied to multiple tasks such as text generation, style control and safety control.

The invention discloses an input and output monitoring system and method for a vibration active control system, and solves the problems that the existing monitoring hardware is complex, the precision is low, the resource occupation is large, and the control performance is influenced. The system comprises an error signal monitoring module, a controller output signal monitoring module and a state monitoring module, a hardware architecture is built by means of a digital processor (DSP), and an error sensor and a controller are respectively connected with corresponding interfaces of the DSP and externally connected with display and storage equipment. The error signal monitoring module comprises a narrowband filtering unit and an exponential weighted accumulation and correction unit, the controller output signal monitoring module comprises an exponential weighted accumulation and correction unit, and the monitoring method is executed according to the steps of initialization, error signal monitoring, output signal monitoring and state observation. According to the invention, extra signal acquisition hardware is not needed, narrow-band filtering and an exponential weighting algorithm are combined, the monitoring precision and the control performance are improved, and the method is suitable for a vibration active control scene.