CN1431769A - Neural network reversal control frequency converter of induction motor and structure method - Google Patents

Neural network reversal control frequency converter of induction motor and structure method Download PDF

Info

Publication number
CN1431769A
CN1431769A CN03112857A CN03112857A CN1431769A CN 1431769 A CN1431769 A CN 1431769A CN 03112857 A CN03112857 A CN 03112857A CN 03112857 A CN03112857 A CN 03112857A CN 1431769 A CN1431769 A CN 1431769A
Authority
CN
China
Prior art keywords
neural network
voltage
input
rotor flux
flux
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN03112857A
Other languages
Chinese (zh)
Other versions
CN1212702C (en
Inventor
戴先中
张兴华
刘国海
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southeast University
Original Assignee
Southeast University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Southeast University filed Critical Southeast University
Priority to CNB031128572A priority Critical patent/CN1212702C/en
Publication of CN1431769A publication Critical patent/CN1431769A/en
Application granted granted Critical
Publication of CN1212702C publication Critical patent/CN1212702C/en
Anticipated expiration legal-status Critical
Expired - Fee Related legal-status Critical Current

Links

Images

Landscapes

  • Control Of Ac Motors In General (AREA)

Abstract

The method is applicable to high performance control for induction motors. The voltage source inverter and the coordinate transformation constitute the extended pressure controlled inventer. The inveter the controlled induction motor 4 and the load form the composite controlled object. The static state neural network and the integrator (S to the power -1) constitute the neural network inverse, which is connected to the point before the composite controlled object so as to form the pseudo-linear system including the rotor magnetic linkage subsystem and the rotate speed subsystem. Based on the method for synthesizing linear systems, the closed-loop controller including the rotor magnetic linkage controller and the rotate speed controller are made.

Description

The Neural network inverse control frequency converter and the building method of induction machine
One. technical field
The present invention is a kind of building method of AC induction motor controller, is applicable to the high performance control of AC induction motor, belongs to the technical field of Electric Drive control appliance.
Two. background technology
At present, the employing frequency converter drives induction alternating current (AC) motor (abbreviation induction machine) thereby constitutes many fields that frequency conversion speed-adjusting system has been widely used in the transmission of former employing DC motor speed-regulating.Frequency converter kind commonly used in the industry has constant voltage and frequency ratio control of conversion device and vector-control frequency converter.The frequency conversion speed-adjusting system that adopts constant voltage and frequency ratio control of conversion device to drive is realized simplyr, but its dynamic and static speed adjusting performance is all very poor.If higher, generally adopt vector-control frequency converter to the dynamic and static performance requirement of governing system.Yet vector control is a kind of stable state decoupling control method, only the decoupling zero of system relation is only establishment when the motor magnetic linkage reaches stable and keeps constant, and because vector control is a kind of based on motor mathematical model and parameter control method, therefore the frequency conversion speed-adjusting system that adopts vector-control frequency converter to drive changed by induction motor parameter and the influence of load disturbance very greatly, the control of governing system is difficult to reach the high-performance of expectation.For overcoming the dependence of existing Control of Induction Motors method to Mathematical Modeling, inductive motor control system changes parameter and the adaptability and the robustness of disturbance from improving in essence, and then every performance index of raising Control of Induction Motors, need to adopt new control technology or new control method, thereby constitute novel frequency converter.
Three. technology contents
1, technical problem
The Neural network inverse control frequency converter and the building method that the purpose of this invention is to provide a kind of induction machine.The frequency converter that adopts this method construct to go out does not rely on the Mathematical Modeling and the parameter thereof of induction machine, can realize dynamic decoupling real between induction machine rotating speed and the rotor flux, thereby has a good dynamic and static control performance, anti-parameter of electric machine variation and load disturbance ability are strong, can improve every control performance index of induction machine effectively, as dynamic responding speed, steady-state tracking precision and parameter robustness and antijamming capability etc.
2, technical scheme
The Neural network inverse control frequency converter and the building method of induction machine of the present invention are at first general Voltage-controlled Current Source inverter, contrary Parker (Park) conversion and contrary Clarke (Clark) conversion to be formed jointly the part of the Voltage-controlled Current Source inverter (the voltage-controlled inverter that is called for short expansion) of expansion as the Neural network inverse control frequency converter of constructing thereafter; And the voltage-controlled inverter of controlled induction machine and load and expansion made as a whole composition composite controlled object; And then adopt static neural network to add integrator s -1Construct the nerve network reverse of composite controlled object, and make nerve network reverse realize the inverse system structure of composite controlled object by the weight coefficient of adjusting neural net; Nerve network reverse is placed before the composite controlled object then, nerve network reverse and composite controlled object are formed pseudo-linear system, pseudo-linear system is equivalent to the second order integral linearity subsystem of two decoupling zeros, be rotor flux subsystem and rotating speed subsystem, realize dynamic decoupling thereby make between the rotor flux of induction machine and the rotating speed; On this basis, adopt a kind of integrated approach of linear system that two decoupling zero second order integration subsystems are designed rotor flux controller and rotational speed governor respectively, and constitute closed loop controller by rotor flux controller and rotational speed governor; The flux observer that adopts electric current rotating speed flux observation model commonly used and Clarke conversion to form obtains the required rotor flux information of magnetic linkage closed-loop control; At last the voltage-controlled inverter of closed loop controller, nerve network reverse and flux observer and expansion is constituted the Neural network inverse control frequency converter jointly and come induction machine is controlled,, realize high performance for induction motor control to obtain the fine control performance index.Wherein:
1. the building method of nerve network reverse is for adding 4 integrator (s with the static neural network with 6 input nodes, 2 output nodes -1) constitute nerve network reverses with 2 input nodes, 2 output nodes.First input of static neural network is first input of nerve network reverse, and it is through first integrator (s -1) be second input of static neural network, be the 3rd input of static neural network again through second integrator; The 4th input of static neural network is second input of nerve network reverse, and it is through the 3rd integrator (s -1) be the 5th input of static neural network, again through the 4th integrator (s -1) be the 6th input of static neural network; The output of static neural network is the output of nerve network reverse.
2. the method for adjustment of the weight coefficient of static neural network is with the stator voltage component With As the input of composite controlled object, to motor speed ω r, i aAnd i bCarry out data sampling, and estimate rotor flux ψ by flux observer r, with the rotational speed omega that obtains rAnd rotor flux ψ rOff-line is asked single order and second dervative, and signal is done standardization processing, the training sample set of composition neural net ψ r, ω r, Be used for static neural network is trained, thus the weight coefficient of definite static neural network.
3. nerve network reverse, closed loop controller, coordinate transform and the flux observer in the Neural network inverse control frequency converter is dsp controller for adopting digital signal processor, realize that by establishment DSP program the Voltage-controlled Current Source inverter is voltage source inverter commonly used.
4. the structure of Neural network inverse control frequency converter is made of jointly the voltage-controlled inverter and the flux observer of closed loop controller, nerve network reverse, expansion, wherein forward path is followed in series to form by the voltage-controlled inverter of closed loop controller, nerve network reverse and expansion, two outputs of closed loop controller are respectively two inputs of nerve network reverse, and two outputs of nerve network reverse are respectively two inputs of the voltage-controlled inverter correspondence of expansion; Feedback path has flux observer, and it is input as stator phase current i a, i bAnd rotational speed omega r, be output as rotor flux ψ rWith rotor flux angle θ.Rotor flux ψ rWith rotational speed omega rFeed back to closed loop controller, form the control of rotor flux closed-loop control and speed closed loop respectively; Rotor flux angle θ is used for the contrary Parker transform operation of coordinate transform.
3, technique effect
Principle of the present invention is contrary by constructing neural network, control will be converted into to the control of this multivariable of induction machine, close coupling, non linear system, correspondingly just closed loop controller can be designed easily the second order integral linearity subsystem of rotor flux and rotating speed.Owing to really realized, thereby can distinguish the effective control that independently realizes induction machine rotating speed and rotor flux, obtained good rotational speed regulation performance to the dynamic decoupling between rotating speed and the rotor flux.And owing to adopted the neural net that does not rely on the controlled device Mathematical Modeling to realize the inverse system function, thereby improved the ability of anti-induction motor parameter variation of Neural network inverse control frequency converter and load disturbance greatly.
The invention has the advantages that:
A. adopt nerve network reverse, the control of this Complex Nonlinear System of induction machine is converted into the control of simple pseudo-linear system, really realized the dynamic decoupling between induction machine rotating speed and the rotor flux.On this basis, by the appropriate design closed loop controller, can obtain good dynamic and static speed adjusting performance.
B. adopt static neural network to add the inverse system that integrator is realized composite controlled object, the constructing neural network inverse frequency transformer is realized the control to induction machine, be completely free of the dependence of traditional Control of Induction Motors method for Mathematical Modeling, reduced induction motor parameter effectively and changed and the influence of load disturbance, improved every performance index of Control of Induction Motors significantly Control of Induction Motors.
The present invention can be used for constructing new type inverter induction machine is carried out high performance control, not only in the drive system that with the AC induction motor is power set, very high using value is arranged, and be in the drive system of power set at the alternating current machine with other type, application prospect also is very wide.
Four. description of drawings
Fig. 1 is made of the structure chart of the voltage-controlled inverter 3 of expansion jointly Voltage-controlled Current Source inverter 1, coordinate transform 2.The coordinate transform 2 that Voltage-controlled Current Source inverter 1 is wherein arranged and form by contrary Parker (Park) conversion and contrary Clarke (Clark) conversion.
Fig. 2 is the structure chart of the composite controlled object 6 that formed by the voltage-controlled inverter 3 of expansion and controlled induction machine 4 and load 5 thereof.Voltage-controlled inverter 3, induction machine 4 and load 5 that expansion is wherein arranged.
Fig. 3 is the complete mathematical model schematic diagram (left figure) and the equivalent mathematical model schematic diagram (right figure) thereof of composite controlled object 6.
Fig. 4 is nerve network reverse 8 and the composite controlled object 6 common schematic diagrames that form pseudo-linear system 9, wherein left side figure is the connection layout of nerve network reverse 8 and composite controlled object 6, right figure comprises rotor flux subsystem 91 and rotating speed subsystem 92 for its isoboles is a pseudo-linear system 9.Nerve network reverse 8 includes 4 integrator (s -1) and a static neural network 7.
The structure chart of the flux observer 11 that Fig. 5 is made up of electric current rotating speed flux observation model 10 and Clarke (Clark) conversion.Electric current rotating speed flux observation model 10 is wherein arranged, Clarke (Clark) conversion and induction machine 4.
Fig. 6 is added to two component of voltage signals that composite controlled object 6 inputs are used to obtain the neural metwork training data.
The structure principle chart that Fig. 7 is simplified by the closed-loop control system that closed loop controller 12 and pseudo-linear system 9 are formed.Wherein pseudo-linear system 9 comprises rotor flux subsystem 91 and rotating speed subsystem 92; Closed loop controller comprises rotor flux controller 121 and rotational speed governor 122.
Fig. 8 is the structure principle chart with the induction motor speed regulation system of Neural network inverse control frequency converter 13 controls.Wherein Neural network inverse control frequency converter 13 is made up of the voltage-controlled inverter 3 of nerve network reverse 8, closed loop controller 12, flux observer 11 and expansion.
Fig. 9 is to be controller with the digital signal processor DSP, and Intelligent Power Module IPM is the hardware structure diagram of a kind of implementation of the present invention of Voltage-controlled Current Source inverter.Dsp controller 15, photoelectric code disk 14, induction machine 4 and load 5 thereof are wherein arranged.
Figure 10 is to be the realization systems soft ware block diagram of the present invention of controller with DSP.
Five. specific embodiments
Embodiment of the present invention are: at first formed the voltage-controlled inverter of expansion jointly by Voltage-controlled Current Source inverter, contrary Parker (Park) conversion and contrary Clarke (Clark) conversion, the voltage-controlled inverter of this expansion will be as a part of whole Neural network inverse control frequency converter.Secondly the voltage-controlled inverter that will expand and induction machine and load thereof are as a composite controlled object, and this composite controlled object equivalence is the quadravalence Differential Equation Model under the rotor flux coordinate system, the vectorial relative rank of system be 2,2}.Adopt the static neural network (static neural network is multitiered network MLN) of 6 input nodes, 2 output nodes to add 4 integrator (s -1) construct the nerve network reverse of composite controlled object.And make nerve network reverse realize the inverse system structure of composite controlled object by each weight coefficient of adjusting static neural network.Nerve network reverse is serially connected in before the composite controlled object, nerve network reverse and composite controlled object synthesize by two second order integration subsystem (s again -2) be the pseudo-linear system that rotor flux subsystem and rotating speed subsystem constitute, thus the control of the nonlinear multivariable systems of a complexity is converted into the control of two simple second-order integration subsystems.Two second order integration subsystems for decoupling zero, adopt a kind of linear system integration method, as proportion integration differentiation PID or POLE PLACEMENT USING etc., make rotor flux controller and rotational speed governor respectively, rotor flux controller and rotational speed governor are formed closed loop controller jointly.Final formation by the voltage-controlled inverter of nerve network reverse, closed loop controller, expansion and flux observer totally 4 Neural network inverse control frequency converters of partly forming comes induction machine is controlled.
Concrete enforcement divides following 9 steps:
1. construct the voltage-controlled inverter of expansion.At first form coordinate transform by contrary Parker (Park) conversion and contrary Clarke (Clark) conversion, afterwards this coordinate transform and Voltage-controlled Current Source inverter commonly used are formed jointly the voltage-controlled inverter of expansion, the voltage-controlled inverter of this expansion is its input (as shown in Figure 1) with two stator voltage components.The voltage-controlled inverter of expansion will be as a part of whole Neural network inverse control frequency converter.
2. formation composite controlled object.Composite controlled object is formed in the voltage-controlled inverter of the expansion that will construct in the first step and controlled induction machine and load thereof jointly, and this composite controlled object is with two stator voltage components
Figure A0311285700071
With
Figure A0311285700072
Be its input (as shown in Figure 2).
By analyze, equivalence and derivation, for the structure of nerve network reverse and learning training provide basis on the method.At first set up the complete Mathematical Modeling of composite controlled object.Then through equivalence shown in Figure 3, the equivalent mathematical model (shown in the right figure of Fig. 3) of composite controlled object be the quadravalence differential equation under the rotor flux coordinate system, its vectorial relative rank be 2,2}.Inverse system through provable this quadravalence Differential Equation Model of deriving exists, and can determine that two inputs of its inverse system are the second dervative of induction electromotor rotor magnetic linkage and the second dervative of rotating speed, and two outputs are respectively two inputs of composite controlled object.Need to prove, this step only provides basis on the method for the structure of following nerve network reverse and learning training, in concrete enforcement of the present invention, and this step, comprise theoretical proof and some corresponding equivalent transformations and derivation etc. that the composite controlled object inverse system is existed, can skip.
4. adopt static neural network to add 4 integrator constructing neural networks contrary (shown in the frame of broken lines of the left figure of Fig. 4).Wherein static neural network adopts 3 layers MLN network, and the input layer number is 6, and the hidden layer node number is 9, and output layer node number is 2, and the hidden neuron activation primitive uses S type hyperbolic tangent function f ( x ) = e 2 x - e - 2 x e 2 x + e - 2 x , The neuron of output layer adopts pure linear function f (x)=x, and x is neuronic input, and the weight coefficient of static neural network will be determined in next step off-line learning.Add 4 integrator s with static neural network then with 6 input nodes, 2 output nodes -1Constitute nerve network reverse (shown in the frame of broken lines of the left figure of Fig. 4), wherein first of static neural network is input as first input of nerve network reverse, and it is through first integrator s -1Being output as second input of static neural network, is the 3rd input of static neural network again through second integrator; The 4th second input that is input as nerve network reverse of static neural network, it is through the 3rd integrator s -1Being output as the 5th input of static neural network, is the 6th input of static neural network again through the 4th integrator.Static neural network is formed nerve network reverse with 4 integrators, and the output of static neural network is exactly the output of nerve network reverse.
5. construct the observer (as shown in Figure 5) of rotor flux.Flux observer is made up of commonly used electric current rotating speed flux observation model and Clarke (Clark) conversion, flux observer be input as stator phase current i a, i bAnd rotational speed omega r, be output as rotor flux angle θ and rotor flux ψ rWherein rotor flux angle θ is used to realize Parker (Park) conversion and contrary Parker (Park) conversion, rotor flux ψ rFeedback quantity as the magnetic linkage closed-loop control.
6. adjust the weight coefficient of static neural network.(a) general's two component of voltages as shown in Figure 6
Figure A0311285700082
With Form with input is added to the input of composite controlled object as shown in Figure 2, gathers the induction machine rotational speed omega with 6 milliseconds sampling periods rWith i a, i b, obtain ψ by flux observer r, and preserve data { ω r, ψ r.(b) rotating speed and rotor flux signal off-line are asked its single order and second dervative respectively, and signal is done standardization processing, the training sample set of composition neural net
Figure A0311285700084
ψ r, ω r, .(c) adopt the error anti-pass BP algorithm that drives quantifier and learning rate changing that static neural network is trained, through 200 training, neural net output mean square error meets the demands less than 0.001, thereby has determined each weight coefficient of static neural network.
7. form rotating speed subsystem and rotor flux subsystem.Constitute nerve network reverse (shown in the frame of broken lines among the left figure of Fig. 4) by the static neural network of having determined each weight coefficient and 4 integrators, nerve network reverse and composite controlled object compose in series pseudo-linear system (shown in the right figure of Fig. 4), this pseudo-linear system is made of jointly a rotating speed subsystem and a rotor flux subsystem, thereby reached the dynamic decoupling between rotating speed and the rotor flux, complicated nonlinear multivariable systems control has been converted into the control of simple two single argument linear systems.
8. make closed loop controller.Rotor flux subsystem and rotating speed subsystem (shown in the right figure of Fig. 4) are made closed loop controller (as shown in Figure 7) respectively.Closed loop controller adopts proportional plus integral plus derivative controller PID, POLE PLACEMENT USING or the most excellent method of quadratic performance in the lineary system theory to design, in the embodiment that the present invention provides, rotor flux controller and rotational speed governor have all been selected proportion differential PD controller for use, and its parameter tuning is PD=100+15s.
9. form the Neural network inverse control frequency converter.The voltage-controlled inverter and the flux observer of nerve network reverse, closed loop controller, expansion are formed Neural network inverse control frequency converter (as shown in big frame of broken lines among Fig. 8) jointly.Can require to adopt different hardware or software to realize according to different control.
Fig. 9 has provided the schematic diagram of a kind of specific embodiment of the present invention, and wherein nerve network reverse, closed-loop control, coordinate transform and flux observer are that dsp controller is realized by software by digital signal processor, and the system program block diagram as shown in figure 10; The Voltage-controlled Current Source inverter adopts Intelligent Power Module IPM to constitute; Controlled induction machine model is Y90S-4, and the parameter of electric machine is P e=1.1kW; U e=220/380V; I e=2.7A; f e=50Hz; n p=2; n e=1400rpm;
According to the above, just can realize the present invention.

Claims (5)

1, a kind of Neural network inverse control frequency converter and building method of induction machine is characterized in that the voltage-controlled inverter (3) that this method is promptly expanded Voltage-controlled Current Source inverter (1), the common Voltage-controlled Current Source inverter of forming expansion of coordinate transform (2); Voltage-controlled inverter (3) by expansion constitutes composite controlled object (6) jointly with induction machine (4) and load (5) thereof; Use static neural network (7) and integrator (s again -1) constitute the nerve network reverse (8) of composite controlled object (6), and make the inverse system structure of nerve network reverse (8) realization composite controlled object (6) by the weight coefficient of adjusting static neural network (7); Then nerve network reverse (8) is connected on composite controlled object (6) before, the equivalence of nerve network reverse (8) and composite controlled object (6) becomes by two pseudo-linear systems (9) of forming of second order integral form rotor flux subsystem (91) and second order integral form rotating speed subsystem (92) independently, again according to the integrated approach of linear system, as proportional plus integral plus derivative controller PID, a kind of in the methods such as POLE PLACEMENT USING or linear quadratic type optimal design, rotor flux subsystem (91) in this pseudo-linear system (9) and rotating speed subsystem (92) are designed rotor flux controller (121) and rotational speed governor (122) respectively, and rotor flux controller (121) and rotational speed governor (122) are formed closed loop controller (12) jointly; The required rotor flux of closed-loop control is provided by flux observer (11), and this flux observer is made up of electric current rotating speed flux observation model (10) and Clarke conversion commonly used; Constitute Neural network inverse control frequency converter (13) jointly with voltage-controlled inverter (3) the order serial connection of closed loop controller (12), nerve network reverse (8), expansion and with flux observer (11) at last, come induction machine is carried out high performance of control.
2, the Neural network inverse control frequency converter and the building method of induction machine according to claim 1, the building method that it is characterized in that nerve network reverse (8) is for adding 4 integrator (s with the static neural network (7) with 6 input nodes, 2 output nodes -1) constitute nerve network reverse (8) with 2 input nodes, 2 output nodes, wherein: first input of static neural network (7) is first input of nerve network reverse (8), and it is through first integrator (s -1) be second input of static neural network (7), be the 3rd input of static neural network (7) again through second integrator; The 4th input of static neural network (7) is second input of nerve network reverse (8), and it is through the 3rd integrator (s -1) be the 5th input of static neural network (7), again through the 4th integrator (s -1) be the 6th input of static neural network (7); The output of static neural network (7) is the output of nerve network reverse (8).
3. the Neural network inverse control frequency converter and the building method of induction machine according to claim 1, the method for adjustment of weight coefficient that it is characterized in that static neural network (7) is for the stator voltage component
Figure A0311285700021
With
Figure A0311285700022
As the input of composite controlled object (6), to motor speed ω r,, i aAnd i bCarry out data sampling, and estimate rotor flux ψ by flux observer (11) r, with the rotational speed omega that obtains rAnd rotor flux ψ rOff-line is asked single order and second dervative, and signal is done standardization processing, forms the training sample set of neural net ψ r,
Figure A0311285700024
ω r, Be used for static neural network (7) is trained, thus the weight coefficient of definite static neural network (7).
4. the Neural network inverse control frequency converter and the building method of induction machine according to claim 1, it is characterized in that nerve network reverse (8), closed loop controller (12), coordinate transform (2) and flux observer (11) in the Neural network inverse control frequency converter (13) are dsp controller (15) for adopting digital signal processor, realize that by establishment DSP program Voltage-controlled Current Source inverter (1) is voltage source inverter commonly used.
5. the Neural network inverse control frequency converter and the building method of induction machine according to claim 1, the structure that it is characterized in that Neural network inverse control frequency converter (13) is by closed loop controller (12), nerve network reverse (8), the voltage-controlled inverter (3) of expansion is formed with flux observer (11) is common, wherein forward path is by closed loop controller (12), the voltage-controlled inverter (3) of nerve network reverse (8) and expansion is followed in series to form, two outputs of closed loop controller (12) are respectively two inputs of nerve network reverse (8), and two outputs of nerve network reverse (8) are respectively two corresponding inputs of voltage-controlled inverter (3) of expansion; Feedback path has flux observer (11), and it is input as stator phase current i a, i bAnd rotational speed omega r, be output as rotor flux ψ rWith rotor flux angle θ.Rotor flux ψ rWith rotational speed omega rFeed back to closed loop controller (12), form the control of rotor flux closed-loop control and speed closed loop respectively; Rotor flux angle θ is used for the contrary Parker transform operation of coordinate transform (2).
CNB031128572A 2003-02-20 2003-02-20 Neural network reversal control frequency converter of induction motor and structure method Expired - Fee Related CN1212702C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CNB031128572A CN1212702C (en) 2003-02-20 2003-02-20 Neural network reversal control frequency converter of induction motor and structure method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CNB031128572A CN1212702C (en) 2003-02-20 2003-02-20 Neural network reversal control frequency converter of induction motor and structure method

Publications (2)

Publication Number Publication Date
CN1431769A true CN1431769A (en) 2003-07-23
CN1212702C CN1212702C (en) 2005-07-27

Family

ID=4790325

Family Applications (1)

Application Number Title Priority Date Filing Date
CNB031128572A Expired - Fee Related CN1212702C (en) 2003-02-20 2003-02-20 Neural network reversal control frequency converter of induction motor and structure method

Country Status (1)

Country Link
CN (1) CN1212702C (en)

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1303753C (en) * 2005-06-07 2007-03-07 南京工业大学 Non-speed sensor inverse controller for AC asynchronous motor
CN100433537C (en) * 2006-03-08 2008-11-12 江苏大学 Method for controlling bearing-less AC asynchronous motor neural network inverse decoupling controller
CN101917150A (en) * 2010-06-24 2010-12-15 江苏大学 Robust controller of permanent magnet synchronous motor based on fuzzy-neural network generalized inverse and construction method thereof
CN102055390A (en) * 2011-01-10 2011-05-11 江苏大学 Construction method for neural network Alpha-order inverse controller of bearing-free brushless DC motor
CN102361429A (en) * 2011-09-13 2012-02-22 江苏大学 Bearing-free asynchronous motor control method based on neural network inverse system theory
CN102437816A (en) * 2011-10-25 2012-05-02 武汉鑫通科创科技发展有限公司 Adaptive motor motion control apparatus based on neural network
CN102647130A (en) * 2012-04-20 2012-08-22 上海电机学院 Permanent magnet synchronous linear motor control method
CN103219936A (en) * 2013-01-16 2013-07-24 镇江市江大科技有限责任公司 Permanent magnet synchronous motor feed-forward decoupling method based on neural network online learning
CN103414432A (en) * 2013-07-12 2013-11-27 江苏大学 Construction method of hybrid electric vehicle belt driven starter generator controller
CN101919156B (en) * 2007-08-29 2013-12-11 罗伯特.博世有限公司 Electrical machine
CN104765929A (en) * 2015-04-17 2015-07-08 浙江工业大学 Nonlinear state decoupling observation method based on multi-time scale recurrent neural network
CN106411224A (en) * 2015-07-31 2017-02-15 发那科株式会社 Machine learning apparatus for learning gain optimization, motor control apparatus equipped with machine learning apparatus, and machine learning method
CN106411210A (en) * 2015-07-31 2017-02-15 发那科株式会社 Motor control apparatus, machine learning apparatus and method therefor
CN110985541A (en) * 2019-11-21 2020-04-10 东南大学 Controller for controlling active magnetic suspension bearing system and control method thereof

Cited By (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1303753C (en) * 2005-06-07 2007-03-07 南京工业大学 Non-speed sensor inverse controller for AC asynchronous motor
CN100433537C (en) * 2006-03-08 2008-11-12 江苏大学 Method for controlling bearing-less AC asynchronous motor neural network inverse decoupling controller
CN101919156B (en) * 2007-08-29 2013-12-11 罗伯特.博世有限公司 Electrical machine
CN101917150A (en) * 2010-06-24 2010-12-15 江苏大学 Robust controller of permanent magnet synchronous motor based on fuzzy-neural network generalized inverse and construction method thereof
CN102055390A (en) * 2011-01-10 2011-05-11 江苏大学 Construction method for neural network Alpha-order inverse controller of bearing-free brushless DC motor
CN102055390B (en) * 2011-01-10 2012-11-07 江苏大学 Construction method for neural network Alpha-order inverse controller of bearing-free brushless DC motor
CN102361429A (en) * 2011-09-13 2012-02-22 江苏大学 Bearing-free asynchronous motor control method based on neural network inverse system theory
CN102437816A (en) * 2011-10-25 2012-05-02 武汉鑫通科创科技发展有限公司 Adaptive motor motion control apparatus based on neural network
CN102437816B (en) * 2011-10-25 2014-05-07 武汉鑫通科创科技发展有限公司 Adaptive motor motion control apparatus based on neural network
CN102647130A (en) * 2012-04-20 2012-08-22 上海电机学院 Permanent magnet synchronous linear motor control method
CN103219936B (en) * 2013-01-16 2015-05-20 镇江市江大科技有限责任公司 Permanent magnet synchronous motor feed-forward decoupling method based on neural network online learning
CN103219936A (en) * 2013-01-16 2013-07-24 镇江市江大科技有限责任公司 Permanent magnet synchronous motor feed-forward decoupling method based on neural network online learning
CN103414432B (en) * 2013-07-12 2015-11-18 江苏大学 A kind of building method of hybrid electric vehicle belt driven starter generator controller
CN103414432A (en) * 2013-07-12 2013-11-27 江苏大学 Construction method of hybrid electric vehicle belt driven starter generator controller
CN104765929B (en) * 2015-04-17 2017-08-11 浙江工业大学 A kind of nonlinear state decoupling observation procedure based on Multiple Time Scales recurrent neural network
CN104765929A (en) * 2015-04-17 2015-07-08 浙江工业大学 Nonlinear state decoupling observation method based on multi-time scale recurrent neural network
CN106411224A (en) * 2015-07-31 2017-02-15 发那科株式会社 Machine learning apparatus for learning gain optimization, motor control apparatus equipped with machine learning apparatus, and machine learning method
CN106411210A (en) * 2015-07-31 2017-02-15 发那科株式会社 Motor control apparatus, machine learning apparatus and method therefor
US9977411B2 (en) 2015-07-31 2018-05-22 Fanuc Corporation Motor control apparatus with magnetic flux controller and machine learning apparatus and method therefor
CN106411210B (en) * 2015-07-31 2018-09-11 发那科株式会社 Control device of electric motor, rote learning devices and methods therefor
US10416618B2 (en) 2015-07-31 2019-09-17 Fanuc Corporation Machine learning apparatus for learning gain optimization, motor control apparatus equipped with machine learning apparatus, and machine learning method
CN106411224B (en) * 2015-07-31 2019-11-15 发那科株式会社 Rote learning device, control device of electric motor and learning by rote
CN110985541A (en) * 2019-11-21 2020-04-10 东南大学 Controller for controlling active magnetic suspension bearing system and control method thereof
CN110985541B (en) * 2019-11-21 2021-06-18 东南大学溧阳研究院 Controller for controlling active magnetic suspension bearing system and control method thereof

Also Published As

Publication number Publication date
CN1212702C (en) 2005-07-27

Similar Documents

Publication Publication Date Title
CN1212702C (en) Neural network reversal control frequency converter of induction motor and structure method
CN1314193C (en) Magnetic levitation switch reluctance motor radial neural network reversed decoupling controller and method for constructing same
Ilic-Spong et al. Instantaneous torque control of electric motor drives
CN100433537C (en) Method for controlling bearing-less AC asynchronous motor neural network inverse decoupling controller
CN101741297B (en) Radial position fuzzy compensation inverse control method and device for bearingless synchronous reluctance motor
CN101299581B (en) Neural network generalized inverse coordination control frequency transformer for two induction machines and construction method thereof
Kotb et al. Speed control and torque ripple minimization of SRM using local unimodal sampling and spotted hyena algorithms based cascaded PID controller
CN101630936B (en) Neural network inverse controller of brushless DC motor and construction method thereof
CN109660160B (en) Switching duty ratio prediction torque control method
Liu et al. Super-twisting sliding-mode observer-based model reference adaptive speed control for PMSM drives
CN110061676B (en) Bearingless permanent magnet synchronous motor controller based on flux linkage observer
CN1216454C (en) Construction method for contradictorily synchronous control frequency-converter of two-phase induction motor neural net
CN1226824C (en) Method for structuring neural network reversal composite controller of variable frequency speed control system with constant ratio of voltage and frequency
CN101299580A (en) Cooperative controller for two-motor vector control frequency control system and construction method thereof
CN1885708A (en) Constructing method for bearingless synchronous reluctance motor feedforward compensation controller
Dasari et al. Modified Luo converter based FOPID controller for torque ripple minimization in BLDC drive system
CN1216455C (en) Method for structuring mervo network inverse synchronous controller of tow-frequenct changing speed regulating motor system
CN104767452A (en) Self-adaptative inverse decoupling control method based on non-linear filters for bearing-free asynchronous motor
CN113541545A (en) Fractional order vector control method and system for permanent magnet synchronous motor
Bouazdia et al. Performance comparison of field oriented control based permanent magnet synchronous motor fed by matrix converter using PI and IP speed controllers
CN113765452B (en) QPR-based electric vehicle system control method
Benbouhenni A new SVM scheme based on ANN controller of a PMSG controlled by DPC strategy
CN115102447A (en) Predictive control method, apparatus and medium for permanent magnet synchronous motor current source type inverter
CN111082711A (en) Brushless direct current motor backstepping method control method and system
Khanh et al. Optimal Fuzzy-PI PMSM Speed Control Using Evolutionary DE Algorithm Implemented on DSP Controller

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
C19 Lapse of patent right due to non-payment of the annual fee
CF01 Termination of patent right due to non-payment of annual fee