120 results about "Error feedback" patented technology

The invention discloses a tunnel or mine water gushing space-time prediction method and system coupled with a hydrodynamic numerical model, and the method comprises the steps: outputting multi-source data based on an identified and verified underground water numerical model, complementing the missing of measured data, quantifying the difference between the permeability characteristics of a fault and a normal stratum, and coupling the difference to a data system, and tunnel or mine excavation space data are merged. And constructing an LSTM-isolated forest-K neighbor regression coupling model, and configuring a multifunctional module to realize multi-scene data co-training. The preprocessed multivariate time series data is divided into a training set and a test set, hidden features are extracted through a coupling model, anomaly detection results are fused, a residual error correction model is synchronously trained, and hyper-parameters and weights are adaptively optimized according to multi-engineering prediction error feedback. And based on the trained coupling model, carrying out synchronous water gushing space-time prediction by adopting a window rolling strategy, and outputting prediction data meeting engineering precision in combination with residual correction. And reliable technical support is provided for safety prevention and control of engineering construction.

The invention provides an automatic vulnerability detection rule generation method based on LLM and static code analysis, and relates to the technical field of network security and software engineering. The method comprises the following steps of: constructing unified structured representation of project code semantics and vulnerability features, classifying taint'source 'and'sink' based on a heuristic rule and LLM (Logical Language Model), and constructing a taint propagation path assisted by a large language model and performing risk assessment; automatically generating a rule based on a'generation-verification-repair 'closed loop mechanism; according to the method, uniform structured representation of code side information and vulnerability side information is constructed, accurate identification of specific project taint sources and sinks and effective paths in multiple environments is achieved, a path risk sorting mechanism is introduced to achieve priority detection of high-risk paths, an automatic rule of error feedback circulation aiming at correction is established, and the accuracy of detection is improved. And the correctness and the performability of the rule grammar are greatly improved, and logic vulnerabilities are reduced, so that the manual debugging and maintenance cost is reduced.

The invention discloses an intelligent stereoscopic warehouse inventory data synchronization method and device and a computer program product, and the method comprises the steps: collecting multi-source data of an intelligent stereoscopic warehouse, and constructing a multivariate data set; embedding the multivariate data set into a nonlinear mapping space, and constructing an embedded state vector for expressing inventory behaviors; a scheduling weight evaluation function is constructed based on the embedded state vector and historical error feedback, the scheduling weight of the inventory operation is calculated through the scheduling weight evaluation function, a decision is made on the synchronization mode of the inventory operation according to the scheduling weight, and the synchronization mode comprises instant synchronization, delay synchronization or merging synchronization; and based on the scheduling weight and historical error feedback, dynamically correcting the response tension coefficient of each dimension in the embedded state vector, and applying the corrected response tension coefficient to subsequent embedded state vector construction to form a feedback closed loop. According to the invention, a powerful guarantee is provided for efficient, accurate and stable operation of the intelligent warehousing system.

The invention relates to the field of control, in particular to a flight attitude control method and system for an unmanned aerial vehicle, and the method comprises the steps: obtaining an expected attitude angle instruction and actual attitude angle feedback of the unmanned aerial vehicle, calculating an attitude angle error, employing a tracking differentiator to calculate a transition process for the expected attitude angle instruction, and enabling a tracking rate factor to achieve the control of the flight attitude of the unmanned aerial vehicle when the instruction is subjected to step change. Based on initial attitude angle error setting, instruction smooth transition is ensured, based on a transition signal output by a tracking differentiator and an estimated value of an expansion state observer, a basic control quantity is generated through nonlinear state error feedback, and control gain is adjusted on line according to a lumped disturbance amplitude. And predicting the lumped disturbance historical sequence by using a grey prediction model, calculating a predicted value at a future moment, and carrying out algebraic summation on the basic control quantity, a compensation component of the lumped disturbance predicted value and a gyroscopic moment decoupling component to generate an attitude control moment signal so as to realize stable flight of the unmanned aerial vehicle.

The invention provides a peristaltic pump control method for a pneumatic mechanical gripper, which relates to the field of intelligent control, and comprises the following steps: constructing a pneumatic conversion module, a peristaltic pump PID control module, a driving execution module, a bending sensing module and an error feedback processing module; the bending angle difference of the mechanical gripper is calculated and subjected to pneumatic conversion to obtain a bending error; the position type PID outputs a control signal, and Kp, Ki and Kd are set on line through a self-adaptive migration optimization algorithm; adjusting and updating individuals by utilizing a direction perception consensus index, a virtual adsorption point set and an adsorption step length; generating a time offset factor correction time sequence based on the historical response memory function; and the corrected signal drives a motor, and precise control over the mechanical gripper is achieved. According to the method, overshoot and steady-state errors are remarkably reduced, response time is shortened, oscillation is suppressed, manual parameter adjustment is reduced, and robustness is improved.

The invention discloses an improved active-disturbance-rejection control method based on error and input delay compensation, relates to the technical field of industrial control, and aims to solve the problem of insufficient input delay and control precision of a high-order inertial system. The method comprises the following steps of: modeling a controlled object into a dynamic system with inertia characteristics, constructing an input delay compensation link and an error-based active-disturbance-rejection link, inputting a control error and a compensation value into an extended state observer (ESO), estimating an error related state quantity and a disturbance tracking value in real time, calculating a control quantity through a state error feedback control law, and compensating the control quantity; and closed-loop accurate adjustment is realized. The invention further comprises an improved second-order active-disturbance-rejection control scheme to adapt to higher precision requirements. The control structure is simplified, control signals and system output are synchronously controlled, the anti-interference capability and the tracking performance are considered, and the method can be widely applied to high-order inertial industrial system control in the fields of chemical engineering, thermal power generation and the like.

The application discloses a tunnel or mine gushing water space-time prediction method and system coupled with a water power numerical model, and the method comprises the following steps: based on the identified and verified underground water numerical model, outputting multi-source data, complementing the missing measured data, quantifying the difference between fault and normal stratum permeability characteristics and coupling to the data system, and incorporating the tunnel or mine excavation space data. An LSTM-isolation forest-K nearest neighbor regression coupled model is constructed, and a multifunctional module is configured to realize common training of multi-scene data. The pretreated multivariate time series data is divided into a training set and a test set, hidden features are extracted through the coupled model, abnormal detection results are fused, a residual correction model is trained synchronously, and the hyperparameters and weights are adaptively optimized according to the multi-project prediction error feedback. Based on the trained coupled model, a window rolling strategy is adopted to carry out synchronous gushing water space-time prediction, residual correction is combined, and prediction data meeting the engineering precision is output. Reliable technical support is provided for engineering construction safety control.

The application discloses a methanol engine control method, related device and methanol engine, and relates to the field of engine systems. By constructing a physical model of an air system, a first coupling influence compensation term representing the deterministic coupling influence between each actuator can be calculated in real time according to an expected value and an actual working condition. By introducing a nonlinear extended state observer, uncertainty factors, such as methanol injection amount deviation, pipeline carbon deposition, external environment changes and the like, which cannot be covered by the physical model, can be estimated in real time, and a second coupling influence compensation term can be generated. Finally, by a linear state error feedback control law, the first coupling influence compensation term and the second coupling influence compensation term are fused to form a dual-channel collaborative compensation architecture of physical model feedforward and nonlinear extended state feedback, and rapid and accurate control of an intake manifold pressure is realized. Through the cooperation of the physical model feedforward and the disturbance observation feedback, the control robustness of the intake manifold pressure and other key parameters is improved.

The invention discloses a federated learning method and system with a communication efficient mechanism in a heterogeneous environment, and the method comprises the steps: carrying out each round of training by a client through employing a gradient tracking algorithm with a momentum mechanism, obtaining the variable quantity of a trained local model matrix parameter, compressing the variable quantity through employing an error feedback mechanism, and uploading the compressed variable quantity to a central server; the central server calculates the average value of the compressed variable quantity uploaded by the participating clients and sends the average value to all the clients, and the clients and the central server calculate a new round of global model based on the average value; and repeating until the global model converges, and completing the training of federal learning. According to the invention, the calculation amount of the client in federal learning can be reduced, and the communication traffic between the client and the central server is reduced, so that the algorithm convergence time is reduced, and the federal learning efficiency is improved.

The invention relates to the field of aviation electrical engineering and real-time simulation testing, and provides a multi-level collaborative verification method and platform for an aircraft power generation system control strategy and a storage medium, and the method comprises the steps: an offline simulation stage: carrying out the preliminary design and verification of the control strategy; in the real-time simulation stage, closed-loop real-time verification and parameter optimization are carried out on the control strategy; in the rapid prototyping stage, optimized control strategy codes are deployed to a rapid prototyping controller and are connected with real or high-precision simulated power generation system hardware, and hardware-in-loop-entity hybrid verification is carried out; wherein a unified performance reference and error feedback chain penetrating three stages of off-line simulation, real-time simulation and rapid prototyping is established, a verification result of each stage is used as an expected reference of the next stage, and error monitoring and backtracking correction are carried out. According to the method, through offline-real-time-prototype three-layer progressive verification and a model precision progressive calibration mechanism, the confidence coefficient and the one-time success rate of final physical verification are greatly improved.

The invention discloses a permanent magnet synchronous motor full-order active-disturbance-rejection control method based on a disturbance differential predictor, and aims to solve the problem of disturbance compensation lag of traditional active-disturbance-rejection control. Comprising the following steps: establishing a d-q coordinate system mathematical model of the permanent magnet synchronous motor and defining total disturbance; designing a full-order expansion state observer, inputting the converted current, voltage and mechanical angular velocity, and outputting a system state variable and a total disturbance observation value; inputting the total disturbance observation value into an original disturbance differential estimator, and outputting a total disturbance first-order differential estimation signal through backward differential discretization; combining the current tracking error, the total disturbance observation value and the differential estimation signal to design a linear state error feedback control law, and generating a control voltage; a rotating speed-current double closed loop is constructed, and inverter driving signals are generated through inverse transformation and SVPWM to drive a motor. According to the invention, the conversion from feedback compensation to feed-forward suppression is realized, the compensation lag is solved, the dynamic performance and robustness are improved, and the structure is clear and easy to realize.

The invention discloses a DisplayPort Lane self-adaptive gain adjustment method based on DPCD (Dedicated Principal Component Diagram) error feedback. According to the invention, in the process of normally displaying the video by the equipment, the background continuously monitors the error feedback information of the link and performs accurate fine adjustment according to the specific channel problem, the picture transmission is not interrupted in the whole process, and the user cannot completely perceive the adjustment action. And for equipment such as a PC VR head-mounted display with extremely high stability requirements, signal fluctuation caused by temperature rise, cable bending or electromagnetic interference can be quietly processed, blurred screen, splash screen or frame loss of pictures can be avoided, and smooth immersion experience can be always kept. Meanwhile, aiming at the independent adjustment capability of a single channel, the method can quickly position and solve the problem of a specific link, is more efficient than traditional full-link retraining, and also can better adapt to complex and changeable actual use environments.

This invention provides an improved sliding mode active disturbance rejection control (ADRC) method and system for an automotive shimmy system based on an electromagnetic linear actuator. Addressing the system's strong nonlinear characteristics and unknown external disturbances, a mathematical model of the automotive shimmy system based on the electromagnetic linear actuator is established. A nonlinear tracking differentiator is constructed to process the desired input signal of the wheel shimmy angle, rationally arrange the transient process, and synchronously extract the differential signal of the desired signal. A continuous and smooth Ifal function is constructed based on trigonometric sine and inverse proportional functions, and a novel extended state observer is designed to estimate the system's internal and external disturbances in real time. A super-helical sliding mode control law based on the hyperbolic tangent function is designed to replace the error feedback control law in the traditional ADRC, outputting a control voltage. This invention effectively estimates and compensates for the impact of vehicle state parameter changes, model uncertainties, and external disturbances on system stability, improving dynamic performance and control accuracy, and exhibiting good robustness and anti-interference capabilities.

The invention discloses a structure deviation compensation-based heterogeneous large model federated fine tuning method and system. The method comprises the following steps of S1, giving a structure deviation compensation-based heterogeneous large model federated fine tuning framework; s2, the server performs feature distillation based on semantic contrast enhancement; s3, the server performs sparse issuing based on the mask; s4, the client performs two-stage local fine tuning based on structural deviation perception; and S5, the client carries out INT8 dynamic quantization uploading with error feedback. On the premise of ensuring the model precision, the effective load of federal communication is greatly reduced.

The invention relates to a code error feedback method combining static analysis and dynamic analysis, and mainly solves the problems of insufficient error feedback and low efficiency in an Online Judge system: reconstructing codes of correct and error versions, including alignment and renaming of variables, and rearranging statements on the premise of not influencing semantics; static and dynamic multi-level alignment is carried out, in the static alignment, an editing script with the minimum cost is calculated through a heuristic editing distance algorithm, an alignment block and a difference list are obtained, in the dynamic alignment, compiling execution is carried out on source codes of two versions, test cases which do not pass operation are dynamically printed through gdb, and execution track information of the test cases is dynamically printed through gdb; error positioning is achieved, data flow backtracking analysis and scoring are carried out on error positions, and error feedback is generated. The method has the characteristics of high execution efficiency, high accuracy and easier understanding of generated feedback, automatic feedback of error codes can be realized, and the learning efficiency of students in programming tasks is improved.

The application discloses a multi-vehicle cooperative transfer method based on a layered architecture, and belongs to the technical field of multi-AGV vehicle cooperative control. The method realizes high-precision, high-synchronism and strong-robustness cooperative work of the multi-vehicles through the architecture of upper-layer global motion planning (overall motion planning and speed decomposition) and lower-layer distributed execution (single-vehicle execution and error feedback closed loop), and based on a maximum-error-first adaptive cooperative control mechanism, the method realizes the fast convergence of the cooperative overall motion with the smallest control cost by selectively adjusting the vehicle with the largest deviation degree through real-time monitoring of cooperative errors of the vehicles.

The invention discloses a noise shaping SAR ADC (Synthetic Aperture Radar Analog to Digital Converter) circuit structure based on error feedback, which relates to the technical field of integrated circuit design and comprises a time sequence control circuit used for generating and coordinating clock signals and control time sequences required by all modules; the analog signal input end is used for receiving an analog signal to be converted and is coupled to the noise shaping SAR ADC module; the noise shaping SAR ADC module is used for carrying out successive approximation type analog-to-digital conversion on the held sampling signal, outputting a digital code, integrating a noise shaping function and extracting a quantization error signal; the error feedback circuit is used for carrying out integration and gain adjustment processing on the quantization error signal and carrying out feedback superposition on the processed error feedback signal; and the digital output end is used for outputting a final digital signal after noise shaping and low-pass filtering. According to the invention, high-precision conversion is realized while the advantages of simplicity and low power consumption of the SAR ADC are maintained, and the applicability of the SAR ADC in high-end application is widened.

The application relates to a multi-model optimal load prediction method and device based on reinforcement learning. The method can automatically select an optimal model or weight combination according to real-time characteristics and historical performance of the model based on a Markov decision mode. The method uses a reinforcement learning decision mechanism to realize dynamic selection and adaptive fusion of the prediction model, thereby overcoming the shortcomings of the prior art in terms of working condition adaptability, prediction accuracy and stability. Meanwhile, through a reward function and an error feedback mechanism, the method can realize continuous optimization of prediction performance. By using the reward function to consider risk-sensitive rewards and switching penalty terms, the method can avoid oscillation caused by frequent model switching, improve model prediction stability, and ensure the accuracy of the results.

The application provides a peristaltic pump control method for a pneumatic mechanical gripper, and relates to the field of intelligent control, and comprises the following steps: constructing a pneumatic conversion module, a peristaltic pump PID control module, a driving execution module, a bending perception module and an error feedback processing module; calculating a bending angle difference of the mechanical gripper and obtaining a bending error through pneumatic conversion; outputting a position type PID control signal, and online setting Kp, Ki and Kd through an adaptive migration optimization algorithm; updating an individual by using a direction perception consensus index, a virtual adsorption point set and an adsorption step length; generating a time offset factor correction time sequence based on a historical response memory function; and driving a motor by using a correction signal to realize accurate control of the mechanical gripper. The method can significantly reduce overshoot and steady-state error, shorten response time, suppress oscillation, reduce manual parameter adjustment and improve robustness.

The invention relates to a multi-model optimal load prediction method and device based on reinforcement learning. According to the method, an optimal model or a weight combination can be automatically selected based on a Markov decision mode according to real-time features and historical expressions of models; according to the method, a reinforcement learning decision mechanism is utilized to realize dynamic selection and adaptive fusion of a prediction model, so that the defects in the aspects of working condition self-adaption, prediction precision and stability in the prior art are overcome. Meanwhile, through a reward function and an error feedback mechanism, continuous optimization of prediction performance can be realized; the risk-sensitive reward and the switching penalty term are considered by using the reward function, so that oscillation caused by frequent model switching can be avoided, the model prediction stability is improved, and the result accuracy is ensured.

The embodiment of the invention relates to a method and device for generating a prompt, electronic equipment and a program product. The method includes generating a first set of candidate cues for the first cue based on the first cue and an error feedback, where the error feedback includes at least one of an error sample and a user feedback. The method further includes determining a second set of candidate cues for the first cue based on the scores of the plurality of candidate cues in the first set of candidate cues. In addition, the method includes determining a second cue from a second set of candidate cues. According to the embodiment of the invention, the group of candidate prompts related to the prompt are generated by introducing the error feedback of the prompt in the practical application, and the group of candidate prompts are evaluated to generate the optimized prompt, so that the experience feeling of using the model by the user according to the optimized prompt is improved.

The invention provides a thermoplastic machine self-adaptive regulation and control method based on dynamic feedback, and relates to the technical field of feedback control, and the method comprises the following steps: obtaining a plurality of operation assemblies of a thermoplastic machine; carrying out collaborative analysis on the relationship among the plurality of operation components to obtain a direct collaborative component and an indirect collaborative component; establishing an integrated feedback control center, and monitoring a plurality of online operation sensing data sets corresponding to the plurality of operation assemblies; decomposing to obtain a direct cooperative sensing data set and an indirect cooperative sensing data set; inputting the direct cooperative sensing data set and the indirect cooperative sensing data set into a double-logic online feedback control model for synchronous error feedback identification, and outputting a direct cooperative error feedback parameter and an indirect cooperative error feedback parameter; and feedback control is carried out on the direct collaboration component and the indirect collaboration component. According to the invention, the technical problem of product defects caused by insufficient synchronism between operation assemblies in a thermoplastic machine system along with the increase of the operation assemblies in the prior art is solved.

The application discloses a kind of open loop hemispherical resonator gyro error identification methods, the vibration energy of the resonator mainly collected as the input of algorithm, and after following processing: 1) by tracking differentiator reduces initial error, effectively solve the contradiction of overshoot and rapidity, reduce external compensation energy, reduce drive voltage consumption;2) by extended state observer to track the variable trend of hemispherical gyro in real time, the degree of dependence on model is low;3) by nonlinear state error feedback control law to eliminate noise strong and robustness strong anti-interference.Solve the technical problems that prior art exists system stability, control precision is not high.

A composite control strategy of a Vienna power factor correction circuit comprises an active-disturbance-rejection voltage control outer loop and an optimized switch sequence model prediction current control inner loop, the active-disturbance-rejection voltage control outer loop designs LTD to arrange a transition process for an output voltage reference value square value to realize rapid tracking, and designs LESO to observe output and total disturbance in real time. An LESF is designed to realize zero static error adjustment through a linear error feedback rate, a switching sequence model is optimized to predict a current control inner ring, redundant vector preprocessing is adopted, a candidate redundant switching sequence is reduced by half according to a redundant vector selection rule, the calculated amount is reduced, a modulation signal is sent to an SVPWM modulation unit, a PWM signal is generated, a switching tube of a main power circuit is driven, and the switching tube of the main power circuit is controlled. And closed-loop control is realized.

A power management integrated circuit (PMIC) may have a loop controller configured to receive a first error signal from a first driver IC having a first driver powered from a supply voltage and configured to drive a first output signal responsive to a first input signal and a first error detector configured to generate the first error signal based between the supply voltage as detected locally to the first driver IC and a first reference voltage associated with the first driver, receive a second error signal from a second driver IC analogous to the first driver IC, and regulate the supply voltage based on the first and second error signals.

The invention discloses an intelligent piezoelectric beam vibration suppression control method based on a fuzzy improved active disturbance rejection controller. Firstly, a kinetic model and a state-space equation of the intelligent piezoelectric beam are established; thirdly, according to the state-space equation of the intelligent piezoelectric beam, constructing a fuzzy improved active-disturbance-rejection controller, including designing a nonlinear Arcfu function based on an arctan function, and constructing an extended state observer and a nonlinear state error feedback control law based on the nonlinear Arcfu function; the fuzzy controller generates a deviation increment of the vibration displacement tracking error and the vibration speed tracking error according to the deviation of the vibration displacement tracking error and the vibration speed tracking error, and dynamically adjusts a control law parameter by using the deviation increment; finally, the vibration displacement of the intelligent piezoelectric beam is obtained in real time, the fuzzy improved active disturbance rejection controller generates system control quantity according to the vibration displacement, the system control quantity acts on the piezoelectric ceramic piece, and active vibration suppression control is conducted on the intelligent piezoelectric beam. According to the method, high-frequency oscillation is avoided, and the vibration suppression control effect is enhanced.

The invention relates to the technical field of wind generating set control, and discloses a wind generating set variable pitch control method based on variable universe fuzzy linear active-disturbance-rejection control, which comprises the following steps: step 1, constructing a linear active-disturbance-rejection controller, including constructing a linear expansion state observer and linear state error feedback; step 2, determining reference parameters of the linear active disturbance rejection controller through a bandwidth setting method, wherein the reference parameters comprise a controller bandwidth reference value, an observer bandwidth reference value and a compensation coefficient reference value; and step 3, constructing a variable universe fuzzy controller, wherein the variable universe fuzzy controller comprises a fuzzy rule table used for reasoning. According to the invention, by constructing a closed-loop collaborative optimization architecture of variable universe fuzzy and linear active disturbance rejection, the core technical bottlenecks of lack of real-time adaptability of control parameters, insufficient anti-disturbance capability under complex wind conditions and limited power fluctuation suppression effect in the prior art are successfully solved.

This invention discloses a differential steering safety stopping control method for vehicles in severe steering failure scenarios. The method establishes a residual steering capability envelope model considering the coupling effect of longitudinal vehicle speed and acceleration / deceleration. Secondly, it constructs a vehicle motion error boundary to cover trajectory deviations caused by model mismatch, external disturbances, and faults. Thirdly, it constructs a safe stopping trajectory and collision avoidance constraints for dynamic or static obstacles. Finally, in the online safe stopping control module, the residual steering capability constraints and robust safety boundaries are introduced into the safe stopping optimization control framework to solve for the optimal safe stopping trajectory and execute control commands. Combined with error feedback control, this ensures that the vehicle trajectory remains within the target trajectory safety boundary. This invention can ensure that, under severe steering failure conditions, the vehicle fully utilizes its remaining differential steering capability to safely move the vehicle out of the active lane and complete the stop, while ensuring kinematic feasibility and robust collision avoidance safety.

According to the active-disturbance-rejection control parameter optimal configuration method for the unmanned aerial vehicle provided by the invention, automatic and intelligent setting of ADRC parameters is realized through an autonomous optimization mechanism of an index triangle optimization algorithm, and dependence on expert experience is completely got rid of. According to the method, a unique two-stage exploration mechanism and a chaos initialization strategy of the ETO algorithm are utilized, a local optimal trap is effectively avoided, and excellent global search capability is shown in collaborative optimization of multiple parameters such as an extended state observer (ESO) parameter and a non-linear error feedback (NLSEF) parameter. According to the method, through the efficient convergence characteristic and the simple structure of the ETO algorithm, the calculation complexity is greatly reduced while the optimization precision is ensured, so that the ADRC parameter online optimization is realized in a resource-limited system. According to the method, adaptive and global optimal configuration of ADRC parameters is successfully realized, and the technical problem that control precision, stability and real-time performance of a traditional method are difficult to consider at the same time is solved.