20results about How to "Improve control robustness" patented technology

The invention provides a trajectory tracking control method and system for a high-load omni-directional AGV, the omni-directional AGV is driven by a plurality of steering wheel mechanisms which are independently steered and driven, and the trajectory tracking control method is characterized by comprising the following steps: acquiring vehicle state information and expected trajectory information of the omni-directional AGV in real time; on the basis of the vehicle state information and the expected track information, core parameters of a model prediction controller are optimized online through a chaos simulated annealing algorithm, and the core parameters comprise a state weight matrix, a control increment weight matrix, a prediction time domain and a control time domain; wherein the model prediction controller performs state prediction based on a longitudinal-transverse decoupling dynamic model adaptive to a high-load working condition; performing rolling horizon optimization through the model prediction controller by utilizing the optimized core parameters, and solving to obtain an optimal control increment; and converting the optimal control increment into a driving instruction for a multi-steering-wheel mechanism, and controlling the omnidirectional AGV to track an expected trajectory.

The invention relates to the technical field of chemical water treatment, in particular to a cloud side-end cooperative intelligent control method and platform for a chemical water treatment black light factory, and the method comprises the steps: constructing a digital twin model fusing equipment topology and operation parameters, and simulating and optimizing the optimal operation state of a system in an unattended and multi-unit cooperative scene; the edge sensing terminal is used for collecting water quality, energy consumption and equipment operation parameters of each process unit under dynamic working condition disturbance in real time, and the actual operation state of the system is accurately recognized through feature extraction and data fusion analysis; by comparing actual and optimal operation states, extracting deviation parameters, performing multi-dimensional analysis, identifying an operation instability mode and key influence factors, generating a self-adaptive cloud side-end cooperative control strategy and driving an intelligent execution device to realize closed-loop regulation, the response capability, the operation stability and the energy efficiency level of the system to complex disturbance are improved, and the energy efficiency of the system is improved. And unmanned, intelligent and collaborative optimization operation of the chemical water treatment process is effectively supported.

This invention discloses a power regulation method and system for an intelligent traction power supply system; including: calculating real-time transfer power, and calculating compensation current in real time by combining power transfer error and negative sequence current; calculating current tracking value based on current inner loop command value and voltage synchronization signal; designing a feedforward-feedback composite controller to calculate feedforward compensation amount and feedback control amount respectively; calculating the total duty cycle of PWM modulation signal based on feedback control amount and feedforward compensation amount, and driving the railway power regulator to output the actual current value; the control method and system provided by this invention have low engineering implementation cost, are suitable for traction substation RPC application scenarios, and are easy to promote.

A data center cooling system control method and device based on a constraint neural network, the method comprising: acquiring thermodynamic parameters of a data center cooling system in real time, and constructing an original database; preprocessing data in the original database to obtain a power consumption dataset and a chip temperature dataset; training a power consumption prediction neural network using the power consumption dataset and a penalty function; training a chip temperature prediction neural network using the chip temperature dataset and the penalty function; the penalty function for temperature prediction adopts an adaptive penalty function, and a penalty factor is introduced when the model predicted temperature is lower than the actual temperature, so that the model predicted temperature is always higher than the actual temperature; and taking the minimum power consumption of the system as an optimization objective, and taking the actual temperature of the chip being less than the upper limit of the chip temperature as a constraint condition, the optimal control parameters under different environmental temperature and humidity and heat load are optimized. The optimization control result can be maintained within the critical temperature of the chip, and the energy saving of the data center is maximized.

The application belongs to the technical field of traffic control systems, and in particular relates to a large language model control traffic signal method for different types of intersections, comprising: obtaining a historical intersection environment set; and performing unified state and action representation to form a first matrix; training a reinforcement learning model of proximal policy optimization based on randomly sampled states in the first matrix, and forming a traffic signal control sequence trajectory based on the first matrix; extracting time sequence information contained in state data of the traffic signal control sequence trajectory through a convolutional neural network to form latent space features; processing the latent space features, actions and rewards of the traffic signal control sequence trajectory through a linear layer respectively to obtain input features; and forming prediction features after operation of a fine-tuned large language model that has been trained; the fine-tuned large language model is updated and optimized through an encoder of a loss function and a double simulation metric learning to be trained, and the prediction features include predicted states, predicted actions and predicted rewards.

The invention relates to the technical field of metallurgical process control, and discloses a molten pool smelting intelligent dynamic control system based on multi-parameter coupling, which comprises a global perception sensor network, a data fusion and state evaluation module and a precise execution control system. According to the system, heat-fluidization multivariable variables of a molten pool are obtained at high frequency through a global perception sensor network, a data fusion and state evaluation module completes prediction-updating through Kalman filtering, abnormal value elimination is carried out through a Mahalanobis distance threshold, working condition drifting is restrained in combination with self-adaptive R / Q, and the state of the molten pool is evaluated. State evaluation is carried out, the middle temperature, the copper matte grade, the liquid level, the change rate and the circulation strength are generated, an LSTM model is adopted for short-time prediction, control suggestions are generated, an amplitude / slope limiting function and a safety interlocking mechanism are matched, a closed-loop execution process is formed, the accurate execution control system converts the control suggestions into actual control actions, and the actual control actions are controlled. And the state of the molten pool is accurately adapted, operation feedback and state monitoring data are transmitted back, and the intelligent control operation efficiency is high.

The application relates to a hyper-redundant robot adaptive control method and system based on double-loop evolutionary reinforcement learning. The method comprises the following steps: acquiring robot data; constructing an outer loop evolutionary evaluation layer and an inner loop behavior execution layer; the outer loop evolutionary layer generates a plurality of candidate weight vectors of a reward function; the inner loop behavior execution layer trains according to the plurality of candidate weight vectors and the robot data and generates a plurality of fitness indexes; the candidate weight vector corresponding to the highest fitness index is retained, and the plurality of candidate weight vectors are dynamically optimized again to generate the plurality of candidate weight vectors for execution training until the training is completed, and the optimal weight ratio is obtained; and the adaptive control of the robot is realized. By introducing the double-loop evolutionary reinforcement learning mechanism, the dependence on the setting of the reward function weight of the artificial experience in the reinforcement learning control process is reduced, the automatic adjustment of the reward parameter and the adaptive optimization of the strategy performance are realized, and therefore the autonomous decision-making ability and the control robustness of the hyper-redundant robot in a limited complex environment are improved.

The application discloses a quantitative acid adding control system for an AGM battery of an automobile, and comprises a data acquisition module for acquiring electrolyte data in an adding process in real time; the electrolyte data comprises an original flow value and an electrolyte temperature. The original flow value and the electrolyte temperature of the electrolyte are acquired, the original flow value is converted into an actual flow value at a standard temperature in real time, and the measurement error caused by temperature change is eliminated; on the other hand, the adsorption amount of a diaphragm is accurately calculated and is brought into deviation calculation, so that the adding deviation caused by the difference in the adsorption characteristics of the diaphragm is compensated; the deviation amount is corrected in real time by the cooperative compensation module, the core deviation problem in the acid adding process of the AGM battery is solved from the root, and the accurate control of quantitative acid adding is realized.

This invention discloses an intelligent control method and system for motorcycle calipers. The method involves real-time acquisition and preprocessing of multi-modal data, including vehicle speed, wheel speed, piston displacement, brake disc temperature, tilt angle, and lateral acceleration, using multiple sensors. Based on the data, braking demand and adhesion coefficient are calculated, and tilt angle and lateral acceleration are used to identify straight-line, curved, or emergency braking scenarios. The data is input into a pre-trained braking control machine learning model to generate a target braking force variation curve. Based on this curve, the drive motor and caliper are controlled for braking, and the model is periodically optimized based on braking feedback and historical data. A component degradation model is used to predict the braking efficiency decay trend, and the braking force is compensated and the model is updated. Closed-loop control is used to adjust the motor current in real-time to follow the target braking force. The brake disc temperature is continuously monitored, and thermal fade compensation is triggered when the temperature exceeds a threshold. This invention achieves adaptive, precise, and reliable intelligent braking for motorcycles.

The application discloses a kind of weak network scene reactive mode online switching control method, device, equipment, medium and program product, belong to the technical field of station collaborative control, including: acquisition and grid point electrical data;Determine comparison parameter based on the grid point electrical data, according to the comparison parameter and switching trigger condition and recovery trigger condition comparison, judge whether to execute switching mode operation;Wherein, when executing switching mode operation, based on preset transition time, calculate final reactive output instruction;The weak network scene reactive mode online switching control method, device, equipment, medium and program product reserve regulation space for power grid, avoid the risk of resonance that long-term constant voltage control can cause, improve the stability of long-term operation of whole station;Reach the trigger condition of voltage related index, namely quickly start constant voltage control mode, give full play to its strong reactive support advantage, effectively inhibit voltage continuous deterioration, help power grid voltage quickly recover stability, guarantee station and power grid operation safety.

The invention discloses an equipment control method and system based on deep learning, and relates to the technical field of compressor control, and the method comprises the steps: obtaining a historical normal sequence and a current operation sequence, and carrying out the preprocessing to obtain a historical feature data set and a current feature data set; calculating the attention weight of each feature under each time scale based on the historical feature data set; dividing the historical feature data set and the current feature data set to obtain a historical fusion window and a current fusion window under each time scale; combining the attention weight to calculate a fusion value and construct a fusion sample, and obtaining a historical fusion sample set and a current fusion sample under each time scale; constructing and training a prediction model to obtain an effective prediction model under each time scale; and calculating a comprehensive valve opening value at the next sampling moment based on the current fusion sample and the effective prediction model. According to the method, the valve opening of the compressor can be accurately predicted, and the high-precision control requirement of the compressor under the complex working condition is met.

The application belongs to the technical field of humanoid robot motion control, and is especially a Multi-Critic humanoid robot sitting-to-standing conversion control model driven by stages, comprising the following steps: step 1: constructing a humanoid robot sitting-to-standing control environment and collecting state information; and step 2: establishing a stage-driven control architecture, and dividing the robot sitting-to-standing process into a guided descent stage, a stable sitting stage and a dynamic standing stage. The application solves the problem of robot stable sitting posture control under complex contact states by decoupling the stage driving and the Multi-Critic function; reduces the optimization conflict of different motion targets by dynamically activating the corresponding control target through stage state determination; improves the stability and practical application ability of complex contact tasks by respectively evaluating the task completion, action continuity and contact stability using independent value networks; and effectively reduces the contact impact, forward instability and high-frequency oscillation problems in the sitting-to-standing process, as compared with the traditional single-Critic unified optimization mode.

The application belongs to the field of mobile robot motion control, and relates to a mobile robot motion control method with dynamic acceleration and deceleration constraints. The method comprises parameter initialization, acceleration and deceleration curve modeling, dynamic curve adjustment and control output constraint. The application introduces a stage model of acceleration and deceleration of in-place rotation control and walking control, dynamically adjusts the acceleration and deceleration curve in combination with external control signals, environment feedback and curvature change of a path, realizes smooth transition and intelligent adjustment of linear velocity and angular velocity of the mobile robot in the path execution process, and significantly improves the path tracking precision, control robustness and intelligence and safety level of the whole motion control system.

The invention discloses a battery charging and discharging control method and system based on an internal health state, and the method comprises the steps: inputting a control quantity at a current moment, an external measurable parameter, and an internal health state estimation vector at a previous moment into a state space model, and processing the control quantity and the internal health state estimation vector to obtain a state space model; obtaining an estimated change rate predicted value of the internal health state at the current moment and an external measurable predicted value at the current moment, and outputting an internal health state estimation vector at the current moment by combining external measurable parameters; constructing a target function by taking control quantity maximization as a target, constructing a constraint condition based on the internal health state estimation vector at the current moment, and solving the target function to obtain a target control quantity at the current moment; and based on the target control quantity, adjusting the actual charging and discharging current as the control quantity of the next moment until the charging and discharging termination condition is reached. According to the invention, the internal health state of the battery can be accurately sensed, and active closed-loop control based on the internal health state is realized.

This invention discloses an adaptive management method and system for intelligent terminal devices in a distributed power distribution network. The method constructs a self-organizing peer-to-peer communication network covering the entire network, enabling neighbor discovery and information sharing among intelligent terminal devices. Each terminal independently calculates control commands based on local electrical parameters and preset multi-objective functions, and optimizes regional control effects through a collaborative mechanism. An adaptive adjustment strategy for the control algorithm is introduced. When an operational anomaly is detected, the terminal can autonomously locate and isolate the fault, and restore power supply based on backup path judgment or microgrid switching. The central management unit periodically conducts capacity assessments and aggregates terminal model parameters through federated learning. The terminal possesses edge autonomy capabilities, maintaining local control and data retention even during communication interruptions. This invention features rapid response, flexible structure, intelligent collaboration, and strong adaptability, making it suitable for intelligent operation and management scenarios in various types of power distribution systems.

This invention belongs to the field of tailings mining technology, specifically relating to a tailings mining control method and system. By constructing a multi-dimensional over-limit criterion based on deviation, vibration spectrum, and load torque, a working condition reconfiguration mechanism is actively triggered. During the freeze period, a response matrix is ​​constructed using steady-state data, and parameter corrections are solved to achieve online updates of interpolation function nodes and slopes. Furthermore, based on the updated mapping relationship, a locally optimal parameter set is generated for multiple devices, and interference risk is dynamically calculated according to the predicted trajectory. If conflicts exist, the area is redistributed according to operational efficiency, and conflict-free paths are generated. This effectively suppresses execution deviations caused by parameter mismatch, avoids multi-machine trajectory interference, ensures stable and coordinated operation of the system under complex disturbance conditions, and significantly improves operational continuity and overall control robustness.

A fast reactive current control method and system for converters based on centralized sampling on the grid side is proposed. Based on the spectral characteristics of composite harmonics, a clustering method is used to obtain the cluster center harmonic frequency bands of multiple frequency bands. The resonant peaks are determined based on the principle that each resonant peak covers two adjacent cluster center harmonic frequency bands. The reduction ratio of the delay time in the repetitive controller is determined based on the frequency of each resonant peak and the fundamental frequency, resulting in an improved fast repetitive controller. The delay element in the feedback branch of the improved fast repetitive controller is decomposed into an integer-order delay element and a fractional-order delay element connected in series. The fractional-order delay element is compensated to obtain a compensated fast repetitive controller. The fast repetitive composite controller obtained by connecting the compensated fast repetitive controller in parallel with a PI controller serves as the improved current inner loop. Based on the voltage outer loop and the improved current inner loop, fast reactive current control is achieved, realizing rapid reactive current compensation.

The application provides a preset time disturbance observation production capacity control method and device of a reentrant manufacturing system, and relates to the technical field of control science and engineering. In the product manufacturing process of driving the reentrant manufacturing system according to the specified product delivery time limit and the specified system delivery output, the system disturbance observer is used to accurately estimate the external disturbance of the system within the preset disturbance estimation time period before the specified product delivery time limit, and the state feedback control mechanism is used to simultaneously consider the system output control function and the external disturbance influence elimination function, so that the corresponding reentrant manufacturing system can respond to market demand within the specified product delivery time limit, complete the product delivery work of the specified system delivery output, effectively eliminate the negative influence of the external disturbance of the system on the system capacity, and improve the system reliability and control robustness of the reentrant manufacturing system.

The invention belongs to the technical field of multi-agent system control, and particularly relates to a tracking control method of a fractional order multi-agent system under full-state constraint, which comprises the following steps: acquiring fractional order multi-agent system data; constructing a leader-follower fractional order multi-agent system model, and defining a system model of each follower; secondly, a fractional order instruction filter is introduced, a subsequently designed virtual controller is preset to serve as input, a filtering signal serves as output, meanwhile, coordinate transformation is executed, and an error compensation mechanism is designed; aiming at unknown nonlinear terms and external disturbance in the model, constructing a fuzzy logic system to carry out approximation, and meanwhile, designing a self-adaptive updating rule; and designing a virtual controller and a final controller by adopting a backstepping method in combination with a barrier Lyapunov function, and carrying out consistency tracking control on the fractional order multi-agent system. Therefore, the problems of insufficient robustness, complexity, explosion and the like of the controller in the prior art are solved.