224 results about "Learning controller" patented technology

The present invention relates to a hearing aid logging data and learning from these data. The hearing aid (10, 100) comprises an input unit (12) converting an acoustic environment to an electric signal; an output unit (16) converting an processed electric signal to a sound pressure; a signal processing unit (14) interconnecting the input and output unit, and generating the processed electric signal from the electric signal according to a setting; a user interface (18) converting user interaction to a control signal thereby controlling the setting; and finally a memory unit (20) comprising a control section storing a set of control parameters associated with the acoustic environment, and a data logger section receiving data from the input unit (12), the signal processing unit (14), and the user interface (18); and wherein said signal processing unit (14) configures the setting according to the set of control parameters and comprises a learning controller adapted to adjust the set of control parameters according to the data in the data logging section.

Monitoring system and method for use with diabetic treatment management. The system includes: a communication interface configured to permit access to stored raw log data, obtained over a certain time, being indicative of glucose measurements, meals consumed and insulin delivery; and a control unit including an unsupervised learning controller configured to receive and process said raw log data and determine at least one global insulin pump setting of basal rate, correction factor, carbohydrate ratio and insulin activity curve parameters. The system may include a processing unit including a first processor for processing measured data indicative of blood glucose level and generating first processed data, a second processor including at least one fuzzy logic module which receives input parameters corresponding to the measured data, the first processed data and a reference data, and processes the data to produce a qualitative output parameter to determine whether any treatment parameter should be modified.

A method for operating a wind energy system is provided comprising the steps of setting the value of an operational parameter of the wind energy system, measuring a yield parameter of the wind energy system and measuring a condition parameter. Further, the method comprises the step of calculating an optimized value of the operational parameter based on historical data and the outcome of the measurements. The method further comprises the step of resetting the operational parameter to the optimized value of the operational parameter wherein the resetting is such that the yield parameter is optimized. Further, a wind energy system is provided having a sensor unit for measuring a yield parameter of the wind energy system, a sensor for measuring a condition parameter, an actuator for adjustment of at least one adjustable part of the wind energy system, and a self-learning controller. The self-learning controller is connected to the sensor unit and the actuator and receives measurement data from the sensor unit. The self-learning controller performs optimization calculations based on the measurement data and sends instruction signals to the actuator based on the outcome of the optimization calculations for the adjustment of the adjustable part of the wind energy system. The instruction signals are such that the yield parameter is optimized.

The present invention discloses a robot trace tracking control method and system. The method comprises the following steps of: the step S100: establishing a dynamical model of an N-degree-of-freedom rigid robotic system; the step S200: according to dynamic characteristics of the robotic system, performing linearization of the dynamical model of the robotic system along an expected trace; and the step S300: taking expected joint angle and joint angular speed of the robotic system as reference input of a robustness adaptive iteration learning controller, taking actual joint angle and joint angular speed of the robotic system as actual input of the controller, generating tracking errors between the actual input and the reference input, and gradually decreasing the tracking errors through iteration calculation of the controller. The robot trace tracking control method and system can perform tracking control of a robot of uncertainty modeling and random disturbance, can improve the rate ofconvergence and the control precision of tracking control and can meet the requirement of the work speed and the precision of the robot.

The invention discloses a machine workshop task scheduling energy-saving optimization system based on reinforcement learning. The machine workshop task scheduling energy-saving optimization system comprises a scheduling target function module, a basic database, a scheduling rule library, a primary scheduling execution module, an energy-saving optimization rule library, an energy-saving optimization module and a scheduling scheme library. Firstly, scheduling targets are set in the scheduling target function module, and primary scheduling is executed in the primary scheduling execution module by selecting scheduling rules from the scheduling rule library and utilizing basic data of the basic database to obtain primary scheduling schemes; then, energy-saving optimization experiences of experts are typed in from an outside-system knowledge expanding port to obtain energy-saving optimization rules; meanwhile, a learning environment is built in the energy-saving optimization module and comprises a workshop energy-consumption module, a processing time module and a scheduling scheme evaluation model; finally, an energy-saving optimization strategy is obtained for the primary scheduling schemes through interaction between a Q learning controller and the learning environment. The machine workshop task scheduling energy-saving optimization system can reduce energy consumption in a machine workshop production process by means of scheduling and has great significance for energy saving and emission reduction of machine workshops.

The invention discloses an iterative learning trajectory tracking control and robust optimization method for a two-dimensional motion mobile robot. The method includes the steps that firstly, a kinetic equation of a two-dimensional motion mobile robot discrete non-linear motion system model is established; a discrete non-linear state space expression is established; a P type open-closed loop iterative learning controller based on the iterative learning control technology is established; then the robust convergence of the established discrete non-linear control system is theoretically analyzed; then parameter item splitting is conducted on control gains of the P type controller, meanwhile, a quadratic performance index function based on controller parameters is designed, and the purpose is to optimize the control parameters; finally, monotone convergence characteristics of output errors and parameter selection conditions generated when a control algorithm acts on a controlled system are analyzed and optimized, and the two-dimensional motion mobile robot can rapidly track an expected motion trajectory at high precision. The method has the advantages that the robust optimization iterative learning controller is suitable for tracking control in an ideal state and suitable for trajectory tracking tasks under the condition that interference exists outside. A designed iterative algorithm is simple and efficient, introduction of a large number of additional parameter variables is not needed, and engineering realization is easy.

An anti-interference iterative learning control method for space manipulator system for capturing non-cooperative targets includes first analyzing and classifying multi-source interference of the space manipulator system in the case of capturing non-cooperative targets in orbit, and establishing a system coupling dynamic model containing the multi-source interference; designing a disturbance observer to estimate and compensate for external disturbance torque and disturbance torque caused by the target non-cooperation, adopting robust H8 control to suppress the random noise of an internal sensor of the space manipulator system with a bounded norm and evaluated errors of the disturbance observer, and using iterative learning control to perform iterative correction on trajectory tracking errors of the system; finally combining the control based on the disturbance observer, the robust H8 control and the iterative learning control, and forming a complete anti-interference iterative learningcontroller, and solving a gain matrix of the disturbance observer and the anti-interference iterative learning controller. The anti-interference iterative learning control method has the advantages of high anti-interference capability and high trajectory tracking control precision, and can be used for in-orbit high-precision operation control of the space manipulator system for capturing the non-cooperative targets.

The invention discloses a periodic adaptive learning control method of an input saturation mechanical arm system. Firstly, an input saturation nonlinear time-delay mechanical arm system model is established; then, a nonlinear time-varying mechanical arm periodic system equation with a Brunovsky standard form is constructed; a tracking controller and a compensatory learning controller are designed; stability and boundedness analysis is carried out according to a composite energy function method, and finally control parameters are selected to achieve real-time adaptive iterative learning torque control over the mechanical arm periodic system. The control method has the advantages of being better in adaptivity, capable of being well used for carrying out compensation on different degrees of saturation input, good in robustness, and capable of being popularized for achieving control over a periodic system without saturation characteristics and and popularized to a nonlinear periodic system without time-delay items; the controllers are easier and more convenient to obtain.

The invention, which belongs to the technical field of artificial intelligence and control, relates to a neural network and enhanced learning algorithm, particularly to an inverted pendulum control method based on a neural network and reinforced learning, thereby carrying out self studying to complete control on an inverted pendulum. The method is characterized in that: step one, obtaining inverted pendulum system model information; step two, obtaining state information of an inverted pendulum and initializing a neural network; step three, carrying out and completing ELM training by using a straining sample SAM; step four, controlling the inverted pendulum by using an enhanced learning controller; step five, updating the training sample and a BP neural network; and step six, checking whether a control result meets a learning termination condition; if not, returning to the step two to carry out circulation continuously; and if so, finishing the algorithm. According to the invention, a problem of easy occurrence of a curse of dimensionality in continuous state space as well as a control problem of a non-linear system having a continuous state can be solved effectively; and the updating speed becomes fast.

The invention discloses a method for suppressing the torque ripple of a permanent magnet synchronous motor (PMSM) based on robust iterative learning control and relates to the technical field of PMSM rotational speed control. In view of the torque ripple of the PMSM and the parameter perturbation and the external load disturbance in a servo system, a robust iterative learning controller is designed by combining adaptive sliding mode control with iterative learning control. The iterative learning control learns an unknown periodic function based on a system state in the system to suppress the system periodic pulsating torque. The sliding mode variable structure control enables a system trajectory to operate along a sliding mode surface to improve the anti-disturbance performance of the system. In view of a sliding mode control buffeting problem, an adaptive law is designed to estimate system disturbance and compensate an estimated value to the controller, thereby weakening sliding-mode buffeting while guaranteeing system robustness. The method, in combination with the advantages of the iterative learning control and the sliding mode control, suppresses the torque ripple of the PMSM while improving the system robustness.

The invention relates to a control method enabling minimum operating time of a batch process in controller asynchronous switching. The method includes the following steps that: step 1, a hybrid state space model is established according to a discrete state space model; step 2, a two-dimensional iterative learning controller is designed, so that a two-dimensional closed-loop hybrid state space model of the batch process can be obtained; and step 3, controller gain, minimum dwell time of a stable stage and maximum dwell time of an unstable stage which enable optimal control performance for a system are calculated. With the control method of the invention adopted, the average dwell time of a stable subsystem of the batch process can be minimum under the premise that optimal control performance of the batch process can be ensured, and unnecessary process operating time can be decreased; and the maximum average dwell time of an unstable subsystem can be obtained, so that ahead-of-time switching measures can be taken, and therefore, system state escape caused by a situation that controller switching lags behind process switching can be decreased; and operating time can be further decreased, time cost can be reduced, and production efficiency can be increased substantially.

The invention discloses an aero-engine control device based on depth Q study and belongs to the technical field of system control and simulation in aviation space navigation boosting theory and engineering. The aero-engine control device control device comprises an acceleration and deceleration restricting module used for outputting physical limit values of all parameters of an aero-engine, and adepth Q learning controller, wherein the depth Q learning controller is used for obtaining the fuel oil flow of the aero-engine through the depth Q learning method according to the control instruction, feedback parameters and the physical limit values of all the parameters of the aero-engine, and controlling the fuel oil input of the aero-engine according to the obtained fuel oil flow. Compared with the prior art, the engine controller is constructed by adopting the depth Q learning method, the learning time of the engine is prolonged, the response speed is increased constantly, and thereforethe response speed of the engine is increased.

The invention provides a self-learning error correction system and method for a numerically-controlled machine tool, which belongs to the field of servo system control. The system provided by the invention comprises a trajectory generation module, a position loop iterative learning controller, a second controller, a motor and a transmission module. The output end of the trajectory generation module is connected with the input end of the position loop iterative learning controller. The input end of the second controller is connected with the output end of the position loop iterative learning controller. The output end of the second controller is connected with the input end of the motor and the transmission module. The system further comprises an outer layer self-learning speed planner. The input end of the outer layer self-learning speed planner is connected with the output end of the motor and the transmission module. The outer layer self-learning speed planner is connected with the trajectory generation module. The system and the method have the advantages that a numerically-controlled system has the self-learning function of error experience accumulation and autonomous error correction in repeated processing; the processing process of the numerically-controlled machine tool becomes the process of automatic learning; and the processing yield is greatly improved.

The invention discloses a method for iterative learning control of the injection speed of a nonlinear injection molding machine system. The method comprises the following steps: firstly, establishing a nonlinear injection molding machine system model taking the injection speed as an output; secondly, constructing a nonlinear state variable dynamic equation of the injection molding machine; thirdly, designing a batch based time-varying forgetting factor iterative learning controller; and finally, analyzing the stability of an iterative control algorithm and the parameter selection of the controller, thereby realizing expected trajectory tracking control on the injection speed of the injection molding machine system under the condition with external disturbance influence. The method has the advantages that the provided P-type iterative learning control algorithm with forgetting factors does not have high requirements on the accuracy of an actual nonlinear system model, and the controller is simple in structure, has certain adaptability and robustness and can be further popularized and applied to other engineering objects including mechanical arms, fermentation processes and the like.

A control device for electric motors, capable of precisely moving one object by using two electric motors based on periodically repeated commands. The control device includes a first learning controller for calculating an amount of correction so that a positional deviation of a first electric motor is minimized, and a second learning controller for calculating an amount of correction so that a positional deviation of a second electric motor is minimized. The first and second learning controllers are independent from each other, and configured to minimize the positional deviation of the corresponding electric motor. The parameters set in the learning controllers, each defining the response of learning control of each electric motor, are equal to each other.

The invention relates to a multi-carrier air-conditioning infrared signal self-learning controller and a method of the controller. The multi-carrier air-conditioning infrared signal self-learning controller comprises an MCU (Microprogrammed Control Unit) module, a plurality of infrared integrated receiving tubes, four input and gates, an EEPROM (Electronically Erasable Programmable Read-Only Memory) storage unit, a serial port communication unit, a key unit and an infrared transmission unit, wherein the four input and gates, the EEPROM storage unit, the serial port communication unit, the key unit and the infrared transmission unit are all connected with the MCU module; and each infrared integrated receiving tube is respectively connected with the four input and gates and the MCU module. The controller has the beneficial effects that the controller is in a sleeping state at ordinary times and is less in power consumption; infrared remote control signals can cause the external interrupt of a main control microprocessor, so that the main control microprocessor is in a self-learning state; and the infrared integrated receiving tubes are integrated, the crosstalk demodulation of a plurality of carrier encoded signals is realized, and learned carrier frequency bands are accurately distinguished, memorized and duplicated to achieve the purpose of remotely learning the multi-carrier air-conditioning infrared remote control signals.