402 results about "Model free" patented technology

A model-free adaptive controller is disclosed, which uses a dynamic artificial neural network with a learning algorithm to control any single-variable or multivariable open-loop stable, controllable, and consistently direct-acting or reverse-acting industrial process without requiring any manual tuning, quantitative knowledge of the process, or process identifiers. The need for process knowledge is avoided by substituting 1 for the actual sensitivity function .differential.y(t) / .differential.u(t) of the process.

The recently introduced theory of Compressive Sensing (CS) enables a new method for signal recovery from incomplete information (a reduced set of “compressive” linear measurements), based on the assumption that the signal is sparse in some dictionary. Such compressive measurement schemes are desirable in practice for reducing the costs of signal acquisition, storage, and processing. However, the current CS framework considers only a certain task (signal recovery) and only in a certain model setting (sparsity).We show that compressive measurements are in fact information scalable, allowing one to answer a broad spectrum of questions about a signal when provided only with a reduced set of compressive measurements. These questions range from complete signal recovery at one extreme down to a simple binary detection decision at the other. (Questions in between include, for example, estimation and classification.) We provide techniques such as a “compressive matched filter” for answering several of these questions given the available measurements, often without needing to first reconstruct the signal. In many cases, these techniques can succeed with far fewer measurements than would be required for full signal recovery, and such techniques can also be computationally more efficient. Based on additional mathematical insight, we discuss information scalable algorithms in several model settings, including sparsity (as in CS), but also in parametric or manifold-based settings and in model-free settings for generic statements of detection, classification, and estimation problems.

A system for, and method of, position estimation of, or tracking, an object. In one embodiment, the system includes: (1) a map generator configured to create, without requiring a propagation model, a mean map for the region and (2) a measurement matching engine associated with the map generator and configured to match received measurements regarding the object to a number of nearest neighboring grid points in the mean map.

The invention relates to a gray model-free control method for a large time lag system. A gray prediction model is combined with model-free control, and complex objects are controlled by adopting a control method of adding the gray prediction model into the model-free control so as to realize the gray model-free control of the large time lag system. Therefore, large time lag can be controlled effectively by the method, and the method has the advantages of small system overshoot, quick stabilization, strong adaptivity and robust performance, and can be applied to the industrial complex large time lag control process.

An effective non-intrusive data-based monitoring method may reduce the cost of control loop performance maintenance by screening and short-listing those control loops or valves that need maintenance. The invention comprises a data-based, model-free, non-invasive method that can automatically detect and quantify stiction present in control valves. The method does not require the implementation of any additional valve travel test or, as commonly known, bump test of the control loop. The method may detect and quantify stiction using controlled variable (pv), controller output (op) and set point (sp) data. It does not require valve positioner (mv) data.

The invention relates to a model-free adaptive optimized control method for a PVC polymerization process, and relates to an optimized control method for the polymerization process. The PVC polymerization process is a complex industrial process characterized by high inertia, large lagging and non-linearity. A model-free adaptive controller algorithm is designed aimed at a common discrete time non-linear system; particle swarm optimization is used to optimize and set key parameters of the controller; a model-free adaptive optimal controller is used to control a polymerization temperature system obtained by least squares identification; and a simulated experiment result shows that the model-free adaptive controller algorithm based on particle swarm optimization can obtain better control performance indexes.

The intelligent panorama creating process based on two fish-eye images includes pre-treatment of fish-eye images, creating spatial model, optimizing the matching parameters and creating panorama. Owing to the theoretically perfect spatial model, model free dimension number for the matching parameter optimizing process reduced from 8 to 4 and the exhaustive search algorithm capable of avoiding trapping in local minimum, the present invention can find fast the optimal matching parameters, and has the advantages of fast matching speed, high matching precision and high image quality. Furthermore,the present invention has automatic pre-treatment of fish-eye images and automatic matching process, so that it may be usd in automatic matching treatment of great amount of fish-eye images.

The invention relates to a fake medicine discrimination and analysis device, a fake medicine analysis system containing the fake medicine discrimination and analysis device and a fake medicine analysis method. The fake medicine analysis method of the invention adopts model-free analysis method, makes an improvement on the basis of the already existing local linear screening method to ensure the method to be more stable and reasonable and combines with major constituent analysis method to form an integral fake medicine discrimination system. The fake medicine discrimination and analysis device, fake medicine analysis system and fake medicine analysis method of the invention only require the spectrum of a compound to be tested and few genuine tablets so as to realize fast and convenient discrimination of the tablets to be tested.

The invention belongs to the field of intelligent robots, particularly relates to a robot motion decision-making method, system and device introducing an emotion regulation and control mechanism, andaims to solve the problems of robot decision-making speed and learning efficiency. The method comprises the following steps: generating a predicted state value of a next moment according to a currentaction variable and a state value by utilizing an environmental perception model; updating state-based on action variables, state values, immediate rewards An action value function network; obtaininga prediction track based on an environmental perception model, calculating a local optimal solution of the prediction track, carrying out differential dynamic programming, and obtaining an optimal decision based on the model; acquiring a model-free decision based on a current state and strategy as well as minimized state-motion functions and based on the state prediction error, the reward prediction error and the average reward value, generating an emotion response signal through an emotion processing computable model, and selecting a path decision according to a threshold value of the signal.The decision-making speed is gradually increased while learning efficiency is ensured.

A method for determining aerodynamic parameters of model free flight tests is used in model free flight tests including a wind tunnel free flight test, an atmospheric free flight test and the like. Under the situation that a model used in wind tunnel free flight or atmospheric free flight moves in a plane, the method of polynomial time is adopted to match the linear displacement of the centroid of the model in the horizontal direction and the vertical direction and the shooting recording observing value, changing along with time, of the angle of pitch of the model, then, the time-dependent changing rule of the linear displacement, the linear velocity and the linear acceleration of the centroid of the model and the angle displacement, the angle velocity and the angle acceleration of the angle of pitch is obtained, and accordingly the time-dependent changing rule of the resistance coefficient, the lift coefficient and the pitch moment coefficient in the process of model flying is obtained. The application conditions and the range of a data processing method of the model free flight tests are expanded, and the test recorded data can be processed in wide application conditions and ranges to obtain the aerodynamic parameters and the movement rule of the model.

A method controls a system for a multiple control steps according to the reference trajectory of a task of the operation to produce an actual trajectory of the system completing the task of the operation. For each control step, a control input to the system is determined using a model predictive control (MPC) having at least one parameter of uncertainty. The method determines a value of a learning cost function of a distance between the reference trajectory and the actual trajectory and determines, using a model free optimization, a value of the parameter of uncertainty reducing the value of the learning cost function. Next, the method determines a set of control inputs for completing the task according to the reference trajectory using the MPC with the updated parameter of uncertainty.

A new multiresolution analysis (wavelet) assisted reinforcement learning (RL) based control strategy that can effectively deal with both multiscale disturbances in processes and the lack of process models. The application of wavelet aided RL based controller represents a paradigm shift in the control of large scale stochastic dynamic systems of which the control problem is a subset. The control strategy is termed a WRL-RbR controller. The WRL-RbR controller is tested on a multiple-input-multiple-output (MIMO) Chemical Mechanical Planarization (CMP) process of wafer fabrication for which process model is available. Results show that the RL controller outperforms EWMA based controllers for low autocorrelation. The new controller also performs quite well for strongly autocorrelated processes for which the EWMA controllers are known to fail. Convergence analysis of the new breed of WRL-RbR controller is presented. Further enhancement of the controller to deal with model free processes and for inputs coming from spatially distributed environments are also addressed.

The invention provides a model-free wavefront distortion correcting system based on a spatial light modulator. The system consist of a polarizing sheet 1, a telescopic system 2, a beam splitter prism 3, a shrinkage device 4, a CCD camera 5, a computer system 6, the spatial light modulator 7, a wavefront controller 8 and a monitoring camera 9. The common CCD camera is adopted to form images; a Strehl ratio of beams is analyzed according to the model-free iteration theory to obtain control signals of optimization parameters required by a wavefront corrector; the performance indexes of the system such as imaging definition, receiving light energy and the like are directly used as objective functions of an optimization algorithm to optimize and acquire the correction effect close to a diffraction limit; and the optimization parameters are calculated in parallel by a computer with a graph processor, the acquired multi-channel control signals are transmitted to the wavefront controller, and then the wavefront controller drives the spatial light modulator to achieve real-time correction of wavefront distortion phase so as to effectively solve the problem of wavefront distortion in atmospheric laser communication.

The invention discloses a servo system position IP controller model-free self-correcting method based on fractional order. The method includes: setting a reference model, exciting a system to obtain an offline database, initiating the related parameters of the system, updating the database according to input and output data newly collected at a current moment, selecting similar data queues according to a similarity criterion function, selecting the order of a fractional order controller, and using particle swarm optimization algorithm to perform self-correcting on fractional-order IP controller parameters. The method has the advantages that the fractional order feature of the system is considered, performance such as overshoot, rising time, error convergence speed and controller parameter variation is appraised comprehensively, the instantaneity of the system is guaranteed by using the similarity criterion function, the method is unaffected by unmodeled dynamics, and the robustness, disturbance rejection ability and control precision of the system are increased.

A new multiresolution analysis (wavelet) assisted reinforcement learning (RL) based control strategy that can effectively deal with both multiscale disturbances in processes and the lack of process models. The application of wavelet aided RL based controller represents a paradigm shift in the control of large scale stochastic dynamic systems of which the control problem is a subset. The control strategy is termed a WRL-RbR controller. The WRL-RbR controller is tested on a multiple-input-multiple-output (MIMO) Chemical Mechanical Planarization (CMP) process of wafer fabrication for which process model is available. Results show that the RL controller outperforms EWMA based controllers for low autocorrelation. The new controller also performs quite well for strongly autocorrelated processes for which the EWMA controllers are known to fail. Convergence analysis of the new breed of WRL-RbR controller is presented. Further enhancement of the controller to deal with model free processes and for inputs coming from spatially distributed environments are also addressed.

The invention provides a biological fermentation adaptive control system and a control method based on neural network inverse. The control method comprises the following steps: constructing the neural network inverse by using a neural network and a time division multiplexing time delay link, enabling the neural network to achieve an inverse system function of a biological fermentation system by adjusting the weight coefficient of the neural network, connecting the neural network inverse and the biological fermentation system in series to form a pseudo-linear compound system, decoupling the pseudo-linear compound system in a linearization mode into two pseudo-linear subsystems with single input and single output, presetting desired signals, according to the control inputs of each pseudo-linear subsystem and the outputs of the biological fermentation system of current time and prior moment, constantly renewing the parameters of a model-free adaptive controller and the control input of each pseudo-linear subsystem of the following moment, enabling the outputs of the biological fermentation system to follow the preset desired signals effectively, thereby achieving the model-free adaptive control of the biological fermentation system, and improving the control performance of a biological fermentation process.

The invention discloses a ship pipeline route optimum design method which comprises the following steps: establishing a pipeline layout space three-dimensional entity model according to the ship structure and the equipment layout parameters; establishing an equipment model free space and generating an STL form file of an equipment model; processing the STL file of the model to obtain the coordinates and vector coordinates of the vertexes of all triangular plates in the file, and determining the layout space range occupied by the equipment; setting pipeline layout space parameters and constructing a layout space mathematic model; analyzing a pipeline system principle diagram, determining the communication relationship among the equipment connection points and establishing a ship equipment communication relation and pipeline diameter information list; carrying out pipeline route planning by utilizing an optimization algorithm; according to the pipeline diameter and the number of the connection points, planning the pipeline route to obtain a preferred resolution by utilizing a pipeline route optimum design method of combining a maze algorithm, an improved non-dominated sorting algorithm and a concerted evolution algorithm; and constructing a ship pipeline system three-dimensional entity model. The ship pipeline route optimum design method has the advantage of effectively improving the design efficiency of the ship pipeline system.

A deep approximation neural network architecture which extrapolates data over unseen states for demand response applications in order to control distribution systems like product distribution systems of which energy distribution systems, e.g. heat or electrical power distribution, are one example. The method is a model-free control technique mainly in the form of Reinforcement Learning (RL) where a controller learns from interaction with the system to be controlled to control product distributions of which energy distribution systems, e.g. heat or electrical power distribution, are one example.

An adaptive process controller drives a process variable to be substantially equivalent to a set point and adapts the controller gain, the controller reset, and / or the controller rate, based on model free adaptation. The adaptive controller combines a controller gain computed from an oscillation index with a controller gain computed from a steady state estimate and that adapts the controller reset / rate by forcing the ratio of two of the controller proportional, integral or derivative terms to be equal to a predetermined value.

The invention relates to an automatic driving system based on relative-entropy deep and inverse reinforcement learning. The system comprises a client, a driving basic data collection sub-system and astorage module, wherein the client displays a driving strategy; the driving basic data collection sub-system collects road information; the storage module is connected with the client and the drivingbasic data collection sub-system and stores the road information collected by the driving basic data collection sub-system. The driving basic data collection sub-system collects the road information and transmits the road information to the client and the storage module; the storage module receives the road information, stores a piece of continuous road information into a historical route, conducts analysis and calculation according to the historical route so as to simulate the driving strategy, and transmits the driving strategy to the client so that a user can select the driving strategy; the client receives the road information and implements automatic driving according to the selection of the user. In the automatic driving system, the relative-entropy deep and inverse reinforcement learning algorithm is adopted, so that automatic driving under the model-free condition is achieved.

The invention relates to a model-free depth-enhanced learning exploration method and a device. The method comprises the following steps of acquiring a characteristic value according to a sample; inputting the characteristic value into a depth-enhanced learning model and processing the characteristic value to obtain an action value; inputting the characteristic value into a counting model to obtainan action counting value; and determining a decision action according to the action value and the action counting value. The actions of different execution times are selected, and the environment return value of each action can be obtained more comprehensively in the exploration process of the depth-enhanced learning. The exploration efficiency is improved.

The invention provides a model-free adaptive control method based on control input saturation. Under the condition of considering control input has position and rate saturation, the invention designs an adaptive neural network constraint controller based on observer technology; and in the design process, a dynamic anti-saturation compensator is provided for adjusting reference preset value in real time to ensure control input does not enter a saturation region. The method is characterized by, to begin with, carrying out model transformation on a common affine nonlinear system through a feedback linearization method; and then, designing a neural network observer and a constraint controller for the transformed system, and providing a dynamic anti-saturation algorithm to adjust a reference set value online to enable the input of the controller to be always within a constraint range.

The present invention provides a PD-type compact-format model-free adaptive navigational speed control method for an ocean vehicle. The method comprises the steps of: issuing an expected navigationalspeed instruction to an ocean vehicle, namely y*(k); employing a compass to measure a current actual navigational speed of a ship, and calculating a navigational speed error absolute value as input ofa PD-type CFDL_MFAC controller; solving the expected instruction by the controller, receiving and executing the expected instruction by an ocean vehicle feed mechanism to change the navigational speed of the ocean vehicle; and measuring the navigational speed of the ocean vehicle at the next moment through the compass carried on the ocean vehicle, performing difference of the navigational speed and an expected navigational speed, obtaining an absolute value, and taking the absolute value as input of the CFDL_MFAC controller at the next moment. The PD-type compact-format model-free adaptive navigational speed control method for the ocean vehicle considers the influence of environmental disturbance on the navigational speed of the ocean vehicle, and repeatedly performs the process mentionedabove until the actual navigational speed of the ocean vehicle is subjected to stable convergence to the expected navigational speed. The PD-type compact-format model-free adaptive navigational speedcontrol method for the ocean vehicle have a prediction effect on dynamic evolution of the navigational speed of the ocean vehicle so as to reduce the overshoot of the navigational speed control system and perform steady-state regulation of time consuming.

A Model-Free Adaptive Quality Variable control system is disclosed for effectively controlling quality variables on-line in closed-loop fashion. It is able to automatically control quality variables under the conditions where there are significant varying time delays and disturbances in the process. Because of its unique capability, the control system is useful for building flexible and adaptive production systems, achieving Six Sigma quality control goals, and fulfilling the on demand manufacturing needs in the new e-commerce environment.

The invention discloses a self-anti-jamming-based model-free self-adaptive AUV control method. The method comprises the steps that firstly, a differential tracker in self-anti-jamming control is usedfor arranging a transition process for input signals and extracting differential signals, then, an expansion state observer adopts uncertain disturbance of the system as total disturbance for performing real-time dynamic estimation and compensation, signals obtained by tracking of the differential tracker are input into a model-free self-adaptive controller, finally, the interference observed by the expansion state observer acts on an output rudder angle of the model-free self-adaptive controller, and finally AUV motion control is achieved. The contradiction between rapidness and super-tonality of a traditional control algorithm is overcome, the anti-jamming capability of the system is greatly improved, implementation is easy, the calculation amount is small, the robustness is high, the control effect for an unknown nonlinear time varying system is obvious, the method can be widely applied to AUV motion control, and the good control effect is achieved.

The invention provides a model-free casting forming machine which comprises a multi-shaft moving system, a workbench and a moving platform system, wherein the multi-shaft moving system at least comprises an X-axis moving system, a Y-axis moving system and a Z-axis moving system; the workbench is located below the multi-axis moving system; the moving platform system is arranged below the workbench; the moving platform system comprises a moving bracket capable of moving to and fro along a direction parallel to the X-axis; a lifting device is arranged on the moving bracket; the lifting device isused for lifting and supporting the workbench, so that the workbench is linked with the moving bracket. According to the model-free casting forming machine provided by the invention, the workbench can be moved without a lifting tool; casting moulds with large-size and complicated cavities can be processed; and environmental pollution is small.

The invention relates to a robot trajectory tracking auto-disturbance rejection control method based on model-free outer loop compensation. The method comprises the following steps that step 1, an omnidirectional mobile robot system dynamics model is established; step 2, an extended state observer is designed according to the dynamics model; step 3, an auto-disturbance rejection controller is designed according to an analytic acceleration control method, wherein the auto-disturbance rejection controller is composed of two parts, one part is used for compensating for the total disturbance of asystem, the other part is used for the trajectory tracking control over a robot, proportional differential feedback is introduced, and the model-free adaptive controller is obtained according to a false Jacobian matrix estimated value to add model-free adaptive control into the extended state observer.

The invention relates to urban expressway entrance ramp traffic flow control, and provides an urban expressway entrance ramp model-free adaptive control method. The method includes: firstly, establishing a controlled process by regarding an entrance ramp flow as a control input quantity and a traffic flow density as an output quantity; detecting the traffic flow density in real time, determining whether traffic flow density data is lost, correcting the entrance ramp flow of the next moment by employing traffic flow information and the entrance ramp flow of the current moment if not, and rapidly realizing control of the traffic flow density as time goes by; when a controller detects data loss, firstly, compensating the traffic flow density data by employing an estimated value, storing the data in a controlled quantity memory, calculating an error of the density, storing the error in an error margin memory, and further correcting a control scheme; and finally, establishing a control lawto calculate the entrance ramp flow, improving the current control scheme through data compensation, and realizing the control of the traffic flow density by employing the corrected control scheme.

The invention discloses an improved MFAC (Model Free Adaptive Control) based data driving energy conservation control system and method of a refrigerating system. The system comprises a constant freezing water supplying temperature control loop and a variational minimization superheat degree control loop, wherein the constant freezing water supplying temperature control loop consists of a PID controller and a transducer; the variational minimization superheat degree control loop comprises a model free adaptive algorithm controller, an electronic expansion valve and an evaporator. The method is characterized in that the frequency of a compressor is adjusted through the PID controller according to the variation of the system load in order to enable constant freezing water supplying temperature, and as a result, the refrigerating capacity fits the heat load; a relation curve between the system load and the minimum stabile superheat degree of the evaporator is acquired through experiment, thus the load of the system can be calculated in real time, and the minimum stable superheat degree corresponding to the load can be calculated through a linear difference method according to the relation curve; the minimum stable superheat degree corresponding to the refrigerating load is treated as the superheat degree setting value in the variational minimization superheat degree control loop; and the superheat degree of the evaporator is controlled by an MFAC control algorithm with retardation time limitation through the electronic expansion valve.

The present invention discloses an adaptive fault tolerant control method with salient pole permanent magnet synchronous motor precise torque output. The method is characterized in that based on a model free adaptive controller, the reference voltage disturbance quantity caused by parameter change of a motor is obtained, on one hand, the reference voltage disturbance quantity is compensated to a control system, a motor system of parameter change is stabilized to a motor system of a nominal parameter, the dynamic performance of a current loop is improved, on the other end, a torque disturbance estimation value is obtained and is compared with a given torque error threshold value, a failure judgment module is formed, when the appearance of a torque tracking failure is judged, a torque command correction amount is outputted through an adaptive fault-tolerant torque regulator and is stacked with a given torque command, a new correction torque command is generated, and the torque control precision of the system is improved. By using a torque feedback loop control structure, the precise torque output under a complex operation condition by an electric vehicle electric drive system can be realized according to the method.