CN106019947A - Servo direct drive pump control hydraulic system wavelet neural network control method - Google Patents
Servo direct drive pump control hydraulic system wavelet neural network control method Download PDFInfo
- Publication number
- CN106019947A CN106019947A CN201610613307.8A CN201610613307A CN106019947A CN 106019947 A CN106019947 A CN 106019947A CN 201610613307 A CN201610613307 A CN 201610613307A CN 106019947 A CN106019947 A CN 106019947A
- Authority
- CN
- China
- Prior art keywords
- hydraulic system
- wavelet
- control
- hydraulic
- servo direct
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
- G05B13/027—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using neural networks only
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
Abstract
The invention discloses a servo direct drive pump control hydraulic system wavelet neural network control method, belonging to the technical field of hydraulic control. A hydraulic control system comprises a wavelet neural network control part, a servo speed control part, and a hydraulic part. The network control part receives a signal from the sensor, analyzes and processes the signal, and carries out adaptive intelligent controlling, and thus the accurate control is achieved. In a traditional hydraulic servo control system, due to a hydraulic natural frequency, hydraulic stiffness and external disturbance force, the hydraulic system has the characteristics of time-varying parameters and strong nonlinearity, thus a precise mathematical model is difficult to establish and the accuracy and safety of a control system are influenced. Through designing a wavelet neural network to control the servo direct drive pump control hydraulic system, a high control precision can be achieved.
Description
Technical field
The present invention relates to a kind of servo direct driving pump control hydraulic system, belong to technical field of hydraulic.
Background technology
Servo direct driving pump control hydraulic technology is the Novel transmission technology of Hydraulic Field, has energy-conservation, efficient, wide range speed control model
Enclose, high reliability, low noise, easily realize the plurality of advantages such as computer numerical control (CNC), have huge in each field of national product
Potential using value.But servo direct driving pump control hydraulic system belongs to volumetric speed control, and contains driver and motor link, makes
Obtain whole system and there is the difficulties such as low-response, control characteristic difference, in traditional hydraulic servo control system, owing to hydraulic pressure is solid
Frequency, hydraulic damping ratio, hydraulic pressure rigidity and external interference power etc. is had to make hydraulic system become parameter time varying, strong nonlinearity, no
Determine the features such as delay, cause being difficult to that controlled device is carried out accurate mathematics mechanism model and describe, and then affect control system
Degree of accuracy and safety.And using the state space description method of modern control theory is the most only that the approximation to real system is retouched
State, certainly exist certain deviation therebetween, it is impossible to meet the industrial process requirement to control accuracy.Fast for electrohydraulic servo system
There is the problem interfered between speed, stability and accuracy, control to be applied to servo direct by wavelet-neural network model
Drive in pump control hydraulic system, it is proposed that a kind of servo direct driving pump control hydraulic system Wavelet Neural Control method.This controlling party
Method has more preferable adaptive ability and control performance than the control strategy such as regulatory PID control and fuzzy control.
Summary of the invention
The present invention is directed to the shortcomings and deficiencies that existing servo direct driving pump control hydraulic system control aspect exists, by design one
Plant wavelet-neural network model, and use steepest descent method that parameter is carried out self-organizing adaptive optimization, make convergence rate more
Hurry up, control accuracy is higher.
For achieving the above object, the technical solution used in the present invention is: a kind of servo direct driving pump control hydraulic system small echo god
Through network control method, including:
The present invention has involved servo direct driving pump control hydraulic system to have permanent magnetism to provide power source with servomotor.According to vector
Control theory sets up the permanent magnetism mathematical model with servomotor voltage Yu motor speed:
The electromagnetic torque of motor, unit;
Load torque, unit;
Q shaft current;Unit;
Q shaft voltage, unit;
Stator inductance is in the equivalent inductance of q axle, unit H;
Stator resistance, unit;
Rotating part converts the rotary inertia on armature spindle, unit
Rotor mechanical separator speed;
The viscous friction coefficient of D motor;
The parameter of electric machine, referred to as torque sensitivity;
Motor parameter, referred to as back emf coefficient.
Set up the mathematical relationship of servomotor rotating speed and hydraulic cylinder displacement
Pump control hydraulic components of system as directed includes hydraulic pump, pipeline, reversal valve and hydraulic cylinder.Hydraulic pump high pressure chest is between hydraulic cylinder
PressureBeing considered as equal, hydraulic cylinder oil return lateral pressure is approximately zero, and the volumetric efficiency curve approximation of hydraulic pump is straight line,
Therefore the internal leakage of hydraulic pump is that heel pressure is directly proportional.By the leadage coefficient of the internal leakage of hydraulic pump with pipeline and hydraulic cylinder it is,。
Obtaining servomotor rotating speed with hydraulic cylinder flow equation is:
The flow of hydraulic pump;
The rotating speed of servomotor;
Pump high pressure chest is to the pressure between hydraulic cylinder;
Pump delivery;
The effective work area of hydraulic cylinder rodless cavity;
Piston movement displacement;
Leadage coefficient;
Hydraulic pump high pressure chest is to the fluid volume between hydraulic cylinder;
Fluid bulk modulus.
Servo direct driving pump control hydraulic system involved in the present invention, owing to hydraulic pump exists leakage and various friction losses,
Torque needed for drive hydraulic system principle should be made up of pressure torque and loss torque two parts:
Order,,
,
Order,,,
== =
The total torque of pump, unit;
The output torque of pump, unit;
The loss torque of pump, unit;
Delivery side of pump pressure, unit;
The parameter of the torque loss of pump;
The density of liquid, unit;
Fluid kinematic viscosity coefficient, unit;
The air gap separation of rotor, unit m;
The total measurement (volume) of hydraulic system, unit;
The bulk modulus of liquid.
Owing to belonging to multifactor, close coupling between output and the input of servo direct driving pump control hydraulic system, nonlinear asking
Topic, is difficult to solve the relation between them with accurate mathematical model, and can well realize by small echo god's neutral net
Meet the Based Intelligent Control of required precision.The input parameter of system has the voltage of motor, electric current, rotor speed;The row of pump
AmountDeng.Output parameter is mainly the outlet pressure of hydraulic pump。
Small echo in the wavelet neural network of the present invention uses Morlet small echo, and its advantage is the Local Property of time-frequency domain
Relatively good, frequency domain energy compares concentration, and frequency alias impact is less.
IfFor Morlet wavelet function,For input layerIndividual input sample,For output layerIndividual defeated
Go out value,For connecting input layerAnd hidden layer nodeWeights,For link hidden layer nodeSave with output layer
PointWeights.Here arrangeIt is the output layer node threshold values,It is hidden layer node threshold values,It isIndividual implicit
The contraction-expansion factor of node layer,It isThe shift factor of individual hidden layer node, then wavelet-neural network model can be described as:
,
In formula,。
If, then:
Wavelet functionThe frame conditions that should meet:
Wherein,It is respectively the bound of framework.Select hereinTime almost tight frame
Morlet wavelet function:
Samples normalization.In servo direct driving pump control hydraulic system, the span of each sample is certain owing to various factors exists
Gap, for preventing from being absorbed in extreme value when training at the flat site of network, sample is normalized in advance.
In formulaFor input sample;For sample after normalization.
Determine network structure.Network input layer node is several is 4, output node number 1, hidden node number 6.As
Shown in Fig. 1.The input of network and output node are respectively with vectorWithRepresent.
I.e.;
Netinit.Error E is taken as 0.001, learning rateTake 0.1, factor of momentumTake 0.07, frequency of training 500 times,
The contraction-expansion factor of random initializtion wavelet function, shift factorWeights are connected with network。
Sample classification.Servo direct driving hydraulic system sample is divided into training sample and test sample, and training sample is used for instructing
Practicing network, test sample is for testing the output accuracy of wavelet neural network.
Control output.Training sample is input to designed neutral net, calculates network and control output, and it is defeated to calculate network
Go out the error E with desired output.
Modified weight.Revise wavelet function parameter and network weight according to error precision E, control output valve and approach expectation
Value.
IfIt isThe of individual patternIndividual desired output, then cost function based on method of least square is expressed as:
In formula,TheIndividual actual output vector;
TheIndividual desired output vector.
Following partial derivative can be obtained by formula:
Factor of momentum is introduced for accelerating convergence of algorithm speed, its iterative formula is:
In formula,For learning rate.
During network weight adjusts, the starting stage, Learning Step selects larger, so that pace of learning is accelerated,
When close to optimum point, learning rate is smaller, it is ensured that weight convergence.
Accompanying drawing explanation
Fig. 1 is wavelet-neural network model figure;
Fig. 2 show the implementing procedure figure of the present invention;
Detailed description of the invention
1. data acquisition.Servo direct driving pump control hydraulic system is carried out data acquisition, including input and output relevant parameter.
In servo direct driving pump control hydraulic system, there is a certain distance due to various factors in the span of each sample, for preventing
During training, the flat site at network is absorbed in extreme value, is normalized sample in advance.
In formulaFor input sample;For sample after normalization.
Sample classification.Servo direct driving hydraulic system sample is divided into training sample and test sample, and training sample is used for instructing
Practicing network, test sample is for testing the output accuracy of wavelet neural network.
2. determine network structure.IfFor Morlet wavelet function,For input layerIndividual input sample,For
The of output layerIndividual output valve,For connecting input layerAnd hidden layer nodeWeights,For link hidden layer
NodeWith output layer nodeWeights.Here arrangeIt is the output layer node threshold values,It it is hidden layer node valve
Value,It isThe contraction-expansion factor of individual hidden layer node,It isThe shift factor of individual hidden layer node, then Wavelet Neural Network
Network model can be described as:
,
In formula,。
If, then:
Wavelet functionThe frame conditions that should meet:
Wherein,It is respectively the bound of framework.Select hereinTime almost tight frame
Morlet wavelet function:
Network input layer node is several is 4, output node number 1, hidden node number 6.As shown in Figure 1.The input of network
With output node respectively with vectorWithRepresent.
I.e.;
3. netinit.Error E is taken as 0.001, learning rateTake 0.1, factor of momentumTake 0.07, frequency of training 500
Secondary, the contraction-expansion factor of random initializtion wavelet function, shift factorWeights are connected with network。
4. network calculations.Training sample is input to designed neutral net, calculates network and control output, and calculate network
Output and the error E of desired output.
5. modified weight.Revise wavelet function parameter and network weight according to error precision E, control output valve and approach expectation
Value.
IfIt isThe of individual patternIndividual desired output, then cost function based on method of least square is expressed as:
In formula,TheIndividual actual output vector;
TheIndividual desired output vector.
Following partial derivative can be obtained by formula:
Factor of momentum is introduced for accelerating convergence of algorithm speed, its iterative formula is:
In formula,For learning rate.
6. calculate error functionIf,, then the 7th step is turned;If, then continue
Adjust network training parameter, until, turn 5 steps.
7. output learning outcome, terminates.
It is illustrated in figure 2 the implementing procedure figure of the present invention.
Claims (7)
1. a servo direct driving pump control hydraulic system Wavelet Neural Control method, it is characterised in that: comprise the following steps: first
First, servo direct driving pump control hydraulic system is carried out data acquisition, including input and output relevant parameter, and data are carried out normalizing
Change processes;Secondly planned network structural parameters, determine the dimension of input layer and output layer, and determine according to certain mathematical model
The dimension of hidden layer and wavelet basis function;Then initialization network parameter, specification error precision, then sample is trained,
Adaptive adjustment weights are carried out, until meeting the control accuracy of demand according to set precision.
A kind of servo direct driving pump control hydraulic system Wavelet Neural Control method the most according to claim 1, its feature
It is: described input layer parameter includes electric moter voltage, electric current, the rotating speed of rotor, pump delivery;Output parameter is delivery side of pump
Pressure.
A kind of servo direct driving pump control hydraulic system Wavelet Neural Control method the most according to claim 1, its feature
It is: be to prevent from being absorbed in extreme value when training at the flat site of network to given data sample, be normalized.
A kind of servo direct driving pump control hydraulic system Wavelet Neural Control method the most according to claim 1, its feature
It is: establish certain mathematical relationship according to the own characteristic of servo direct driving pump control hydraulic system, devise input and output layer
Dimension, it is determined that the dimension of hidden layer, its network schemer is 461 types.
A kind of servo direct driving pump control hydraulic system Wavelet Neural Control method the most according to claim 1, its feature
Being: propose control method based on wavelet-neural network model, wavelet neural network is with wavelet basis function as neuron
The feed-forward network model of excitation function, the product that it is Wavelet Analysis Theory and neutral net be combined with each other, there is wavelet transformation
Good time frequency localization feature, and neutral net has the ability of the strongest generalization ability and self-adapting data, therefore small echo
Neutral net has higher self-organizing, self study and an adaptive ability, and approach, fault-tolerant and inferential capability.
6. the network designed by is 3 layers of BP neutral net, and the wavelet basis function of hidden layer is used Morlet wavelet function, its table
Reaching formula is。
A kind of servo direct driving pump control hydraulic system Wavelet Neural Control method the most according to claim 1, its feature
It is: sample is used classification to use steepest descent method to be trained after processing by described wavelet neural network, until error precision
Reach requirement.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610613307.8A CN106019947A (en) | 2016-07-31 | 2016-07-31 | Servo direct drive pump control hydraulic system wavelet neural network control method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610613307.8A CN106019947A (en) | 2016-07-31 | 2016-07-31 | Servo direct drive pump control hydraulic system wavelet neural network control method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106019947A true CN106019947A (en) | 2016-10-12 |
Family
ID=57115330
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610613307.8A Pending CN106019947A (en) | 2016-07-31 | 2016-07-31 | Servo direct drive pump control hydraulic system wavelet neural network control method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106019947A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109083887A (en) * | 2018-09-13 | 2018-12-25 | 西安建筑科技大学 | A kind of method for diagnosing faults of the excavator hydraulic system based on ACA-BP algorithm |
CN109657789A (en) * | 2018-12-06 | 2019-04-19 | 重庆大学 | Gear case of blower failure trend prediction method based on wavelet neural network |
CN111507012A (en) * | 2020-04-26 | 2020-08-07 | 太原科技大学 | Method for establishing gas dissolution theoretical mathematical model of rolling shear hydraulic system |
CN111796511A (en) * | 2020-07-24 | 2020-10-20 | 华北电力大学 | Wavelet neural network PID (proportion integration differentiation) online control method and system of hydraulic actuator |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH10327530A (en) * | 1997-05-23 | 1998-12-08 | Mitsubishi Electric Corp | Ratio-differential relay for protecting transformer |
US20030018599A1 (en) * | 2001-04-23 | 2003-01-23 | Weeks Michael C. | Embedding a wavelet transform within a neural network |
CN102710214A (en) * | 2012-06-11 | 2012-10-03 | 桂林电子科技大学 | Nonlinear interference control method and control system for permanent magnet linear synchronous motor |
CN104636800A (en) * | 2013-11-06 | 2015-05-20 | 上海思控电气设备有限公司 | Freezing station system neural network optimization unit based on least square weighted and method thereof |
CN104989633A (en) * | 2015-06-04 | 2015-10-21 | 中国航空工业集团公司上海航空测控技术研究所 | Aircraft hydraulic pump fault diagnosis method based on bionic wavelet transform |
CN105634356A (en) * | 2016-01-07 | 2016-06-01 | 江苏大学 | PLC-based generalized neural network inverse internal model implementation method for multi-motor speed regulating system |
-
2016
- 2016-07-31 CN CN201610613307.8A patent/CN106019947A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH10327530A (en) * | 1997-05-23 | 1998-12-08 | Mitsubishi Electric Corp | Ratio-differential relay for protecting transformer |
US20030018599A1 (en) * | 2001-04-23 | 2003-01-23 | Weeks Michael C. | Embedding a wavelet transform within a neural network |
CN102710214A (en) * | 2012-06-11 | 2012-10-03 | 桂林电子科技大学 | Nonlinear interference control method and control system for permanent magnet linear synchronous motor |
CN104636800A (en) * | 2013-11-06 | 2015-05-20 | 上海思控电气设备有限公司 | Freezing station system neural network optimization unit based on least square weighted and method thereof |
CN104989633A (en) * | 2015-06-04 | 2015-10-21 | 中国航空工业集团公司上海航空测控技术研究所 | Aircraft hydraulic pump fault diagnosis method based on bionic wavelet transform |
CN105634356A (en) * | 2016-01-07 | 2016-06-01 | 江苏大学 | PLC-based generalized neural network inverse internal model implementation method for multi-motor speed regulating system |
Non-Patent Citations (9)
Title |
---|
FAA-JENG LIN 等: "Adaptive Wavelet Neural Network Control for Linear Synchronous Motor Servo Drive", 《IEEE TRANSACTIONS ON MAGNETICS》 * |
WANG HONGYAN 等: "A Hybrid Adaptive Wavelet Neural Network Control and Sliding Mode Control for Electro-hydraulic Servo System", 《PROCEEDINGS OF THE 29TH CHINESE CONTROL CONFERENCE》 * |
WANG HONGYAN 等: "Robust Output Feedback Control of Electro- Hydraulic Servo System using Wavelet Neural Network", 《2009 CHINESE CONTROL AND DECISION CONFERENCE》 * |
吴建新 等: "电液伺服系统的神经网络在线学习补偿自适应控制", 《液压气动与密封》 * |
张世英 等: "《协整理论与波动模型 金融时间序列分析及应用》", 31 May 2009 * |
李笑 等: "液压泵恒流控制的神经网络实现方法及应用研究", 《中国机械工程》 * |
武志勇 等: "小波神经网络在液压装置压力测试中的应用", 《电子测量技术》 * |
王新 等: "《机电设备故障诊断技术及应用》", 30 April 2014 * |
陈艳: "小波神经网络在平整机液压控制系统的应用", 《液压与气动》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109083887A (en) * | 2018-09-13 | 2018-12-25 | 西安建筑科技大学 | A kind of method for diagnosing faults of the excavator hydraulic system based on ACA-BP algorithm |
CN109083887B (en) * | 2018-09-13 | 2020-04-21 | 西安建筑科技大学 | Fault diagnosis method of excavator hydraulic system based on ACA-BP algorithm |
CN109657789A (en) * | 2018-12-06 | 2019-04-19 | 重庆大学 | Gear case of blower failure trend prediction method based on wavelet neural network |
CN111507012A (en) * | 2020-04-26 | 2020-08-07 | 太原科技大学 | Method for establishing gas dissolution theoretical mathematical model of rolling shear hydraulic system |
CN111507012B (en) * | 2020-04-26 | 2023-04-14 | 太原科技大学 | Method for establishing gas dissolution theoretical mathematical model of rolling shear hydraulic system |
CN111796511A (en) * | 2020-07-24 | 2020-10-20 | 华北电力大学 | Wavelet neural network PID (proportion integration differentiation) online control method and system of hydraulic actuator |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Ho et al. | Design and control of a closed-loop hydraulic energy-regenerative system | |
CN106019947A (en) | Servo direct drive pump control hydraulic system wavelet neural network control method | |
Sarkar et al. | GA-optimized feedforward-PID tracking control for a rugged electrohydraulic system design | |
Li et al. | Counterbalancing speed control for hydrostatic drive heavy vehicle under long down-slope | |
CN102790581B (en) | Constructing method for robust controller for radial position of bearingless asynchronous motor | |
CN106681152A (en) | Method of using PSO optimized trap parameters for servo resonance suppression | |
CN103941635B (en) | A kind of system and method suppressing drilling rod stick slip vibration | |
Zad et al. | Robust Model Predictive position Control of direct drive electro-hydraulic servo system | |
Lee et al. | Design and experimental evaluation of a robust position controller for an electrohydrostatic actuator using adaptive antiwindup sliding mode scheme | |
Shen et al. | Adaptive fuzzy sliding mode control based on pi-sigma fuzzy neutral network for hydraulic hybrid control system using new hydraulic transformer | |
CN102384119B (en) | Method, device and system for speed regulation control of electro-hydraulic proportional valve and engineering machinery equipment | |
Yu et al. | Design, mathematical modeling and force control for electro-hydraulic servo system with pump-valve compound drive | |
Ahn et al. | Self tuning fuzzy PID control for hydraulic load simulator | |
Wei et al. | Speed control strategy for pump-motor hydraulic transmission subsystem in hydro-mechanical continuously variable transmission | |
Yahya et al. | Research on simulation analysis of valveless hydraulic servo system | |
Haifeng et al. | Modeling and simulation of EHA system based on fuzzy adaptive PID control | |
CN102790582B (en) | Manufacturing method of radial fuzzy neural network generalized inverse controller of bearingless asynchronous motor | |
Pratumsuwan et al. | Force and position control in the electro-hydraulic system by using a MIMO fuzzy controller | |
Wei et al. | Self-tuning dead-zone compensation fuzzy logic controller for a switched-reluctance-motor direct-drive hydraulic press | |
Shen et al. | Speed tracking control for hydraulic transformer system based on active regulating common pressure rail | |
Yu | Intelligent neural network control strategy of hydraulic system driven by servo motor | |
Ali et al. | H∞ LOOP SHAPING CONTROL DESIGN OF THE ROTATIONAL VELOCITY OF A HYDRAULIC MOTOR | |
Jiangtao et al. | Mathematical modeling and fuzzy adaptive PID control of erection mechanism | |
CN105216777B (en) | Ratio pressure brake control valve | |
Al Mhdawi et al. | Adaptive PI-Fuzzy Like Control of a Stack Pneumatic Actuators Testbed for Multi-Configuration Small Scale Soft Robotics |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20161012 |